Turtle,
It is nice to see some of the right questions being asked when approaching this as an optimisation problem.
I would posit that the changing environment (i.e. rules and condition changes between comps) rules out a static location of minima.
An argument could also be made that the selection process is not deterministic, the fittest x team/cars are not always the top x team/cars at comp.
I also think that our human idea of randomness (usually a uniform distribution) is not what we should expect. It doesn't matter if large proportion of a population are incredibly similar, and we should expect this. At any given time the FS field should have a lot of common traits, and we would expect when new successful information enters the population that other teams will absorb it quite rapidly. What would help in the long run is some ability to encourage the creation of new information. There are plenty of ways to achieve this algorithmically, and good ways to do it in practice in FS, especially given that points can be directly given for innovation.
...
Claude,
I wouldn't expect to see an amazing design report any time soon. I have been involved in FS since 2001, both as a student and as a Faculty advisor (Including 4 design event wins, and a number of placings). In that time there has been almost no feedback given to the teams I have been involved with as to the structure and quality of the design report. There have been exceptions including from you in Oz 2012, and Pat's advice to improve drawings (given to all teams).
Without a decent feedback mechanism there is little possibility for teams to significantly improve their reports.
I know of a few teams trying to share some information here, including the team I advise. On ECU's facebook page the team has released 3 past design reports, the most recent being 2012 (A report you provided some feedback on). These posts have been the amongst the best viewed posts the team has put up.
Our students would love to have a better insight into the processes other teams use in the design of their vehicles. However apart from a few teams that release good reports, or are quite willing to share (i.e. GFR, Monash, UQ, and others). This seems like such a small amount of available information for decades of work conducted largely at the expense of publicly funded institutions.
If teams are so against releasing the top of a given year how about releasing all reports 2-3 years after a given competition has ended? (Although I like your idea as well)
Heck we could treat it like most information developed at universities, and put it behind a paywall, that in no way benefits the authors. At least the information would be available.
Kev