Frankly, I'm probably not good enough nor have the time to devote to something that could be quite large, but I will be keeping up with this.
I plan on testing myself at the given parameters just to see if my times are "quick" or not.
Cool beans.
Also in our 3rd season Championship we introduced success ballast of 70, 50 and 30kg for 1st, 2nd and 3rd place drivers with it being halved if finish outside podium on next race then removed if outside again, which evened the championship out.
Part of my experiment is to see if a ballast like you're mentioning can be worked out for races who are not part of a championship and therefore have no knowledge of their opponents skill level.
My Trial Run
In Arcade Mode I took the Mclaren MP4-12C for 5 laps around the Nordschleife Nürburgring with all aids off and track grip set to 'real'. ABS was set to 1 and tyres were chosen as SH although the test drivers will be using SM.
I had never driven this car before so was experiencing it for the first time and was rather blown away by it's speed and abilities. I drove 5 laps and then exited to get my lap times. My laps were all valid, no red times, although I'm not the fastest driver out there and my laps were a bit cautious so they aren't to be taken as 'good' times.
Here are the times I recorded from my five consecutive laps, no more, no less.
1 - 7. 45. 999
2 - 7. 42. 243
3 - 7. 42. 412
4 - 7. 43. 119
5 - 7. 42. 727
I rounded up or down these times^ to the nearest second and got these times.
1 - 7. 46.
2 - 7. 42.
3 - 7. 42.
4 - 7. 43.
5 - 7. 43.
I then added these times together and got
37. 16. I then divided this number by the number of laps in this particular test, five. The product of this calculation was
7. 432. Rounding this up to the nearest .5 I made
7.5.
Once I have completed the other test combos I will do the same procedure to calculate an average and then using that data I can workout how a system can be implemented to include a ballast.
I will note at this point that the test runs can be done as many times as a driver wants but must only be done in sessions of the number of allowed laps per combo. e.g run 5 MP4-12C/ Nurburgring laps, collect times, see times aren't so good, run another 5 laps.
By the way, we're still looking for drivers so if you'd like to join the experiment drop a post in here!