OK, here's the story...
Newly "partial" rebuilt EA82T. Built on a solid core. Professionally reconditioned gen1 heads. All new gaskets etc. New H20 pump, oil pump, etc. Nothing left to chance.
Maybe a little more history on the "solid core", was it ever severely over heated or left out in freezing weather without antifreeze?... I'm thinking maybe cracked block? Not common on Subies but always possible, and could get water and oil mixed without effecting the compression.
Or compression trouble in #4 (HG or rings, assuming the head rebuild didn’t miss any cracks), and just enough water in the cylinder to pump the numbers up to match the other holes. Try a leak down test, or drain the water (don’t forget the plugs in the heads) and retry the compression test.
Side note: I cant think of any way you could get higher than normal compression from a valve problem, at least not at cranking speed.
Assume nothing, check everything, just some thoughts