Sort of, but I don't like to file on a dropout. I know it's expedient and gets you down the road, but I'm a snob, so I think a dropout that's been filed that way is a Fail.

Yes the properly-dished wheel is the final arbiter of whether the two blades are the same length. Centering under the crown is a must. The square can tell you that too, but the wheel is more sensitive, i.e.any discrepancy at the square will be magnified up at the crown, so you can be more precise, using just you eyeballs, with the wheel.

If the wheel is off-center after getting the tips equidistant from the center plane, I prefer to add rake on the long side to shorten it, or take some curve out on the shorter blade. Which of those two options I choose depends on the existing bend situation; sometimes it doesn't matter but sometimes there's one blade that should get bent

If you started with both tips the same fore-aft, then after the rake changes, one tip will be forward of the other. So I'll usually bend whichever blade got the rake adjustment, up near the crown. If I decreased the curve, that tip is back compared to the other, so I'll bang that one forward such that it bends in the heat-affected zone near the crown. The amount of bending is so tiny you will never see it, no riples or cracks in the paint. I feel I need to state that, because I've had some people say bending the fork there is unsafe. Maybe it is, and I've been doing it wrong since the '70s? But zero fork failures (that I know of!) on forks I aligned in all that time, says any added danger is pretty theoretical. I'll accept that theoretical danger to not have to file on the dropout <ugh!>

Mark B