I'm going just by the bedroom counts listed on that website. It's possible that those numbers are also deceptive because square footage per bedroom is sufficiently different. I was mainly remarking that the breakdown seems to indicate that NY inventory is biased towards 2BR apartments while SF inventory is biased towards 1BR. This is based merely on the fact that the SF overall average is much closer to the SF 1BR average while the NY overall average is much closer to the NY 2BR average.

Price per square foot would shed more light on this, but I don't know of a reliable (free) source of that data.