It is an interesting intellectual concept, but as pointed out by @rbscebu in a previous post, it was originally used to make a reference surface not a finished part. A machinist lapping a part will use a lapping plate to do so, not a reference surface. Essentially a substrate commonly impregnated with an abrasive. The lapping plate or part would periodically be checked against the reference surface for flatness but never lapped against it.If only using two, and one is steel with diamond attached, how is it going to get convex or concave. That is good enough for me.
Paper on glass, the same, to an acceptable degree for honing razors.
The three stones is an interesting intellectual process, but really, just use one flat surface that doesn't wear, to flatten the one that does.
In our world, the part is a hone, the lapping plate is a diamond plate, tile and wd, glass and SiC, etc and the reference surface could be a straight edge.
I guess that was a really long winded, pontificating way to say i completely agree with you but it is early and i have not had any coffee yet.
Also a side note, for the 3 plate method to be completely effective the parts need to be matched on at least 2 axis. Not really effective on a rectangle as you can not get full contact when your bottom stone is laying north south and your top stone is laying east west.