Susan, it seems to me that water's first two priorities are; 1) to seek it's own level or the lowest place it can get and, 2) Evaporate, ie - leave the place it's worked so hard to get (somewhere further down it's agenda is to give life, esp. to algae) What I've noticed on the many pools I've worked on is that a hole in the cover is the prime suspect, if you have one and try to pump the water off your cover, you will lose pool water (however, the faster you pump off the cover, the less pool water you'll lose). I can't question that some water may make it through a 'holeless' cover, but I think the amount would need to be measured with a well calibrated micrometer.
So I offer this other scenario:
When you get water from precipitation on top of the cover the overall level of the pool water is raised (both above and below the cover). Water in a pool freezes from the top down. Only the top ~10% of an iceburg is above water. If the water on top of the cover freezes, it will try to 'sink' into the pool, forcing the still liquid water up and possibly over the top edge of the pool. Since water expands when it freezes, when it melts you will finally notice that you've lost some water. (This isn't as elloquent as I might like, but should do for starters) In short, don't worry too much about the 3 missing " of H2O, until you notice a 'larger than expected' drop in the pool water after running it.