I regularly try to compare gear and am finding it very difficult. Normally I'll just swap pieces and then mouse over whatever skills I'm using to look at the new damage, but socketed items are problematic. I've done pretty extensive testing with gems and most gems don't do anything to the average (paper) damage of the skills they should be influencing. Is this a display bug or is it also manifesting itself in game play? An easy way to replicate my testing is to grab a weapon with sockets and look at your average damage as well as damage of spells. Sometimes slotting a crit gem changes the spell's average damage but most of the time it doesn't.

Am I looking at this wrong or is there some way to accurately compare different pieces of gear for optimal itemization?