The big trick to using any collimator (or "bore-sighter," even if they don't use the bore for alignment) is to first put it on several rifles that are already sighted-in. Anybody who does this will soon notice two things:
1) Even sighted-in rifles don't always show the scope's reticle in the same place on the collimator's grid. This is partly because of differences in scope height over the bore, but mostly because different barrels vibrate differently. So there isn't any universal "sight-in point" even for the best collimators.
2) The average reticle placement for various rifles often won't be in the middle of the collimator's grid. This can occur with both spud and magnetic collimators, but often the average will be close enough to get the scope reasonably aligned--if you adjust the scope so the reticle's on the "sweet spot," rather than the center of the collimator grid.
I've had one of the Leupold magnetic collimators since they were first offered, and it's always worked pretty well. An acquaintance got one and it didn't work at all, and thought I was lying about mine. I offered to prove how well mine worked, but (like a lot of people, including some on the Campfire) he was convinced only his experience was valid. But the fact is there'll always be some variation, often considerable, in "affordable" products.
Some collimators can be adjusted so they line up better. My primary collimator is an old Bushnell with three expanding spuds, and I adjusted it to line up years ago--but even then the first shot will be up to eight inches from point of aim at 100 yards, again because different rifle barrels vibrate differently.
The most valuable use of a reasonably accurate collimator is in mounting scopes. With the scope's adjustments centered, if the reticle lines up reasonably well with the bore the scope will not "run our of adjustment" when we attempt to sight in in. PLUS the adjustments will be more accurate, because the internal mechanics will line up correctly. Many scope "malfunctions" are actually due to crooked mounting, and a correctly used collimator will help prevent that.
+1
My main purpose for grid-type collimators is two-fold:
1. I've noticed, as MD stated above, that most of my zero'd rifles are in approximately the same place on the collimator, but not every rifle. HOWEVER, for any given rifle and mount setup, I zero the scope and then take note of where the crosshair lands in the collimator. I file that away in my notes. Then if I want to confirm zero or mount another scope on that rifle/mount platform, I can quickly reference my notes and make sure the reticle is set to the same place in the collimator. This method has gotten new scopes to within 0.1-0.2 MRAD so far. I use the Bushnell magnetic unit, FWIW. And even new rifles and scopes are within about 1 MRAD of intended POI most of the time, when set to the average zeroed location in the collimator.
2. This has become an invaluable tool for preliminary tracking tests. I use the reticle of my scope to measure the distance between lines in the grid of the collimator (confirmed with several different scopes). Then I dial from one hash mark to another, and make sure the subtended hash marks match the amount of adjustment dialed into the turret. I do this over 10, 15, or even 20 MRAD if the reticle allows it, and get a pretty good sense of not only the adjustment increments, but also whether the scope tracks plumb or not, and if there are any dead spots or oddities in the adjustment range.