I really wouldn't put much stock in it.
Axis Games played: 984
Allied Games played: 1851
Allied winrate is higher because allies are playing more games and thus the chance of playing against players outside the top 150 is far more likely.
There's two ways to do this properly:
Expand the bracket to 1000-1500 instead of just top 150.
Only include games played between players in the top 150. (Probably not a large enough sample size if you do this though, and it's not possible without being Reric.)
While I partly agree on team games (3vs3+), this assumption doesn't work out in 1vs1 games as all the ladders are separated, hence the ladder with the most players should've the highest winning rate, as we have most players competing there. Yet this is not the case and balance problems in 1vs1 amplify in team games, that's why 2vs2 is in even a worse state.
In 1vs1 you just wait longer to get a proper ranked opponent as allies, it doesn't mean you get more noobs to bash - rather less.