If the piston top is parallel to the crank(along the same axis as the wrist pin and crankshaft) a mis-alignment of the cylinder bore for which a half gasket would compensate will show itself as differential squish. The holes in the side of my block where the crankshaft roller bearing carriers set, were not bored parallel to the deck. But my piston top was parallel to the crankshaft(machinists level can help you check this also). So when I installed the cylinder without a half gasket, this angular difference showed up in the differential squish. The difference in squish across the 4" piston equaled the angular offset between my crank and deck. That told me the cylinder top and bottom were at least parallel, but not if the installed cylinder formed a parallellogram or a rectangle. It also created a wicked offset of the piston on the wrist pin from crank up to crank down: As 38ac mentioned, that is the important part to align the cylinder to, either by using a half gasket to align the cylinder bore or by relocating the crank bearings to align the piston axis to the deck and cylinder. The point I am getting at is that if the piston to rod offset is small from full up to full down, you are probably going to have a pretty small differential squish across he piston top over the wristpin ends.
The cylinder/liner true can be checked fairly easilly on a large drill press. I put two bolts into the drill press table slots that would fit inside the cylinder when it is placed onto the table over the bolts. I would have them touch the inside of the cylinder around the 3 and 6 Oclock positions. I clamp a rod with a dial indicator into the chuck and position it down inside the top of the cylinder with it's measurement rod touching the cylinder top between but directly above the two bolts. You rotate the cylinder slowly keeping the bottom inside edge of the cylinder against the two bolts. As you rotate the cylinder watch the dial indicator. If the cylinder bore is perpendicular with the lower surface, the dial indicator measurement should stay constant thru 360 degrees of rotation. If it varies thru 360 degrees, something is out of true. I have a pic somewhere of this setup, I will see if I can locate it and post it...