This is all very interesting. Glad I was bored and found this yesterday
I ended up doing some research of my own, and found some interesting stuff. I always use Nicola Acciarri's 2004 physics and magic data for the original tracks cause I've already got the hang of it and have great setups for all tracks. This way, playing the game is really a breeze...if I want to race I just load my setup, go straight to qualifying and then race. Anyway, after looking through the magic data I found out that all his data already used tyre 55 for Hards and Softs. I never thought about it, but I assumed the softs had better performance with more wear, but then I decided to try it out and here's my results:
2004 physics and magic data by Nicola Acciarri
820 Bhp Perf File - 790 Bhp hofxlap
Hard Tyres - 4.0 fuel - Barcelona 2003
1st lap - 1:14.630
2nd lap - 1:14.895
Soft Tyres - 4.0 fuel - Barcelona 2003
1st lap - 1:14.909
2nd lap - 1:14.847
Now I don't mean to brag, but I'm a fairly consistent driver so it means that either tyre gives the same performance and wear. Which is good, since it means the game doesn't mess with the tyre properties after set in the track data.
At this point I had already changed the fuel weight and consumption as described by Ali and decided to run a 17 lap stint to check what would happen:
01 - 1:17.698
02 - 1:17.385
03 - 1:17.293
04 - 1:17.086
05 - 1:17.119
06 - 1:16.943
07 - 1:16.705
08 - 1:16.770
09 - 1:16.612
10 - 1:16.891
11 - 1:16.279
12 - 1:16.442
13 - 1:16.185
14 - 1:15.861
15 - 1:15.868
16 - 1:15.467
I then compared that to a previous 2003 Spanish GP I did a few days ago, and noted that in that occasion, my lap times were all around 1:16.500 +/- 0.4s for the first stint. So, indeed Ali's suggestion seems good. I'll now make a new race in Barcelona to check what happens with the AI and I'll also change the soft tyre to 54 and see how that fairs in comparison to 55.