There are a number of competing factors that are in conflict regarding the classes in endurance racing. Entertainment may limit innovation, cost controls are not exciting, BoP is always contoversial, waivers piss everyone off. I don't pretend that my ideas are the answer, but here they are.
In order to maintain a high diversity of cars and a sufficiently large number of teams competing in the class, it required that the car makers produce a road car that would be competitive within the rules. The perceived need to keep the racing close for "entertainment" purposes brings pressure to adjust the performance of this diverse set of cars to some magic lap time that all can achieve, or failing that to use other gimmicks that "equalize" the performance over a race distance, like modifying fuel capacity or fuel flow that would affect pit stop strategy. At least the organizers haven't stooped to "push-to-pass" or DRS.
Another factor that affects the class is allowing cars that would not normally fall within the regulations to compete through the granting of a waiver. The Ford GT did not meet the homologation requirement of producing the minimum number of road cars prior to entering competition, so they were granted a waiver. The Viper was granted a waiver in GTLM in the IMSA series even though the displacement of the engine exceeded the maximum displacement in the regulations by 2.5L. There are many other examples of this practice.
Car makers want to win races to advertise their cars as "better" than their competitors. "Win on Sunday sell on Monday". This aspect places great pressure on the organizers to "level the playing field" to allow any of the cars to win. To a similar extent teams also want the chance to win to attract sponsorship and drivers as well as satisfy their competitive drive. If someone actually made a better car, shouldn't they be allowed to race it without being hobbled? In his racing autobiography Mark Donohue stressed that what his team (Penske) always strove for was the "Unfair Advantage", in fact that is what the bio was entitled. I submit that all teams strive for such an advantage and that impulse advances the sport.
One of the current concerns is that the newest entrants to the class: Ford GT, Ferrari 488, and BMW M6 are dominating the class over the older entrants and the pressure to conform has resulted in charges of "sandbagging" and the issuance of BoP adjustments between races and even between qualifying and the race. These moves are controversial to say the least. Some are concerned that the "gaps" between the classes will be eroded if these new cars aren't slowed down. Could there be a better answer?
Maybe. I think the entire class structure might benefit from a re-evaluation. The following are just my thoughts on one way that might help solve this puzzle of what to do with the endurance classes. There are probably many others that could answer as well.
I would like to see a prototype class that meets some definition of "road legal" but otherwise ignores minimum production numbers. This "GT Prototype" class would contain the halo cars not only for cars like the Ford GT, but would allow early entry for the mid-engined Corvette and the next Aston Martin that are currently on the drawing board and headed for competition in the near future. Such a class would keep the high cost of these puppies away from the real modified production models that typify GTE/GTLM and still allow the factory teams a place to play. This should be a real prototype class: eventually the cars in this class would show up in some form as a production model. This would be different from the current definition of "prototype" in LMP1 and LMP2 which will never be produced in road-going form. At some point this "GTP" class might supplant LMP2.
The successor to the current GTE/GTLM class would be for real road cars produced in numbers and the racing parts that make them competitive would be required to show up as parts available through dealerships. This approach was used in the original Trans-Am regulations (1966 - 1973), and parts that were developed for racing had to be made available to the public (even if "for off road use only"). This should make for easier entry into the class for new teams. Cost caps on the performance parts might need to be imposed in such a scheme. Regulations would be strictly enforced and changes/adjustments would happen only between seasons. Cars must conform to their homologation forms, which is where any needed "balancing" would take place; allowed modifications would be spelled out similar to the way the SCCA handles their varied classes.
There is currently a divergence in the "lower" GT classes between IMSA and ACO/FIA: the IMSA GTD class is basically GT3 with Pro/Am driving teams, which relies entirely on BoP while the ACO/FIA runs GTE-Am which basically runs the same cars as GTE-Pro but with Pro/Am driver lineups (as well as eliminating the newest spec cars). Both of these approaches have their pros and cons and either could accomplish the goal of bringing in more car and driver diversity. I favor the GTE-Am approach, but there is something to be said for cars that have a place in more than one series.
The five current Prototype classes need a re-think in my opinion. LMP1/H is now stricly a WEC beast. They are for factory works teams only and it is practically mandated to be the overall winning class. Why is there this fiction that the LMP1-privateer class is part of LMP1? They should be classified separately from LMP1/H. LMP2 evolved from LMP675; a lightweight that could compete against their heavier and more powerful LMP1 (evolved from LMP900) cousins; until the LMP2s started winning some important races overall, then suddenly it was thought neccessary to cost cap the class and make them heavier than the LMP1 cars. Now the ACO/FIA are requiring a spec engine and only four chassis options as well as Pro/Am driver teams. LMP3 has a similarly restrictive formula, intended to replace the aging spec LMPC cars.
In my proposed scheme I would like to see a lightweight prototype that would be allowed to compete for overall wins, but against the cubic euros spent in LMP1/H that is very unlikely to ever happen. Barring that, LMP2 should still be less restrictive: allow more chassis and engine suppliers and impose a cost cap if needed. Requiring production-based engines would make the class more relevant to road cars, but if race engines prove more cost effective over the course of a season, then perhaps that restriction could be dropped. If LMP2 can't compete for overall wins, then there are only two purposes for them: a less expensive way to go prototype racing, and allowing non-professional drivers a place to go prototype racing. Keeping it Pro/Am (like the ACO/FIA) is probably justified.
In IMSA the proposed DPI serves a different purpose that LMP2 in the ACO/FIA; it will be the top class, expected to take overall wins, so they should be able to run all-Pro drivers. They should also be able to use any tire they want. I would drop the requirement to run only one of four chassis and restricting bodywork to single chassis/engine combinations. This bodywork restriction is aping the old Daytona Prototype model which came about belatedly in the Grand-Am series and has resulted in only one manufacturer bodykit. Allowing bespoke bodywork should be an option, but not a requirement and any engine should be usable in any chassis. A more generous cost cap might be in order. My feeling is that DPI, modified along the lines proposed here, could form the basis for a replacement for LMP1-privateer. If that happened then there could be a place for DPI at Le Mans without upsetting the Pro/Am LMP2 teams.
If there is any place for a spec formula in endurance racing, then it would be in the replacement for the LMPC class. Currently the thinking is that this would be LMP3, which isn't quite spec. I have mixed feelings about this class. A fully spec chassis/engine/tire class would make for a better driver training prototype entry platform, while normally a spec car is anathema to me personally. As it is currently, it is neither fish nor fowl. LMPC worked quite well for a long time, so maybe the full-spec formula would work well as a replacment too. At least there would be a ready market for used cars in this class, making entry even more affordable.
In summary my imaginary class structure would look like this (names could be changed, but these refer to existing classes where applicable):
LMP1 - Leave it as is and let the ACO/FIA do with it what it will. This is a WEC-only class.
LMP1-privateer/DPI - loosen the DPI regulations and converge these two classes into one, classified separately from LMP1
LMP2 - Pro/Am prototype run by privateers only with cost caps. Could become Pro/Am versions of DPI (maybe)
LMP3 - Full spec Pro/Am prototype. Training ground and low-cost prototype entry platform.
GTP - Road legal prototype halo cars. Works teams with Pro drivers. The playground for dream cars.
GTE-Pro - True modified production cars with strict regs and "factory" parts. Pro drivers.
GTE-Am - GTE cars with Pro/Am drivers. Alternatively GT3 cars would form this class
If you think that 7 classes are too many, recall that the 1966 Le Mans 24 had 9 classes. It is only in more recent times that it was thought that 4 was the most we should have. The only place all 7 would run is in the WEC. There would probably be a natural selection process that would eliminate unneeded classes. Each series would choose which classes they support.
The goal of such a re-organization is to reflect reality, eliminate the BoP controversies as much as possible, stimulate fan interest, and try to keep everyone happy. Getting the organizing bodies, the manufacturers, the teams, and the fans to agree on something so radical is probably not possible. These are my thoughts, at present they are only valuable to me.