*Is this true? That you are compared against those who take it the same day? What if you take the test at 8:00 AM vs. 4:00 pm?

A smaller sample size could be skewed either way. Say a fewer number of people took it, but all were high achievers. That would hurt you. Of course the converse is true, but I would think enough people are taking it every day that it averages out.

The percentile figures are determined in advance - for example, you know going in that if you score a 760, it's 99th percentile, even if everyone else in the world who takes the GMAT that day scores a 790. The GMAC has enough data from previous tests to know what the percentile breakdowns are.

(Note, though, that the percentiles are adjusted every so often; for example, a 750 used to be 99th percentile, while it is "only" 98th today.)

So, there's no difference between a weekday and a weekend, other than whichever works best with your schedule and "mental clock."