Usually 20ish stages per GT, so divide overall by winner by 20. Round 2.5 up to 3pts for Tour stages and Giro/Vuelta 2 pts per stage. I don't think 2 stage wins should be equal to a podium place on at the biggest race on the calendar. Distribution of pts for non podium placings at Giro/Vuelta should be lower. No way should a stage win in a minor race be equal to a top 10 in the Tour. There is never going to be a perfect system, but it should be focused primarily on the major races, not minor races or stage wins. The big guns don't care about the minor races.Time for some self ctiticism:
After reading all the comments I agree that there are some flaws in the system. Stage wins are the biggest problem. Maybe 1/10th of the GC win would be fair:
5 points for Tour stage
4 points for Giro stage
2 points for stage in Paris-Nice etc.
1 point for stage in smaller races.
Lower places in the GC should also be counted:
Tour top 10 GC: 50/25/10/8/6/5/4/3/2/1
Giro/Vuelta top 10 GC: 40/20/8/6/5/4/3/2/1/1
Paris-Nice top 3 GC: 20/10/5
That way the big sprinters would lose a few places, and climbers would get some extra points.
However I'm not going to do all that counting. An update for active riders once a year would be easy, but doing it thoroughly for the whole cycling history would be a lot of work. The biggest changes would be in the lower part of the ranking.
About Bernal: I was also surprised. He's currently at 165 points, so he's likely to make it next year.
I am still laughing at the idea of LeMond and Fignon giving up their Tour victories for a load of sprint stages, comedy gold.