Doesn't that mean you are assuming that all 1000 and 500 are seeking concurrently? RL doesn't work that way. The raw number of games played doesn't say anything about when they were played or what the Allied skill level was like when it was played.
I was simplifying it a bit to make my point clear. In reality we assume that both factions have an equal chance of playing better skilled, equally skilled and lesser skilled players. However if there's 1000 axis games in the top 150 and 500 in the allied top 150, that means that there's a lot more axis games in the lesser skilled player category than the better skilled/equally skilled. It's simply impossible otherwise: axis always have to play against allies.
The difference between a top 150 player and a top 300 player is enough to account for a 10% difference in win rate.
I just think that using this site as any sort of evidence of balance is stupid. It can help indicate trends, but it's not absolute. The reality is that OKW playing quite a bit differently will throw players off at a high level of play. And they will lose more games to OKW because of it. Likewise, Brits being changed is enough to make people who had a very simplistic playstyle before struggle because the things they relied on were changed. Most players are strict metagame followers and don't have an adaptive bone in their body.
I know the logic behind it, it's stupid. 
Dullahan
Strategos
Wall $treet Fat Cat
Street Bit
pensss
Relic Entertainment
