Backs can be hard to understand, much more so than tops, and particularly Classical tops.
What I find with ladder braced backs is that the lowest mode is often the fundamental bending mode along the back. There will be two node lines, roughly across the wide part of the lower bout, and between the waist and the shoulders. These may be fairly straight or slightly curved.
The next mode up is often the lowest torsion mode, with one line along the length in the center and another across the the plate, so it's divided into quadrants. Sometimes these will be swapped in frequency with the torsion mode lower in pitch, as it would normally be on an unbraced plate. The cross braces add some stiffness in twisting, and raise the pitch.
As you go up from there in pitch you will alternate between something that is predominantly a lenthwise bending mode and one that combines bending and twisting. Thus the third mode often has two crosswise node lines, and the fourth a lengthwise line with two crosswise ones, and so on.
As you go up in pitch crosswise bending will also become more important. This manifests itself as curvature in the node lines. At some point the curves will start to close in on themselves, so you'll get 'rings' of one sort or another.
Although it is undoubtedly true that the 'free' plate modes could be used to predict the modes of the completed instrument, it's also the case that the line from one to the other is neither obvious nor simple. These are complex systems, and although they are not so nonlinear as to be chaotic, it's still true that small differences in initial conditions can diverge fairly rapidly.
What this means is that you can learn to read the modes of the backs you build, and unlitmately gain some ability to get things to come out where you want. However, since your methods and work differ from mine, my experinece may not help much in predicting what your instruments will do, especiallly in terms of frequencies.
I would say, though, that in my experience it's a lot more effective to shave the lower crooss braces than the upper ones. When I went back to using ladder braced backs some years ago I found I was removing a lot of wood from those (generally four brace backs), to the point where the lower braces looked pretty flimsy. The backs were also lower in pitch than I wanted. I went to using low, wide braces for the lower two, wich reduuces their stiffness to weight ratio, so that I end up with 'beefier' lower braces and a slightlly heavier back than I would have had I started out with taller, narrower ones. I don't feel that added weight is as much a penalty on a back, and may even help. Now that I've been at it for a while I find I can usually get things pretty close to where I want them, although, this being wood and all, there are always exceptions.
One last thing: hold the speaker in an active area, and place the pads on node lines. If the pads are even a little off the lines will not form well, and the frequency will be different from the 'real' mode pitch. Again, with practice you get to where you know what you're looking at, and can usually get the pads pretty close to where you'll want them and know where to hold the speaker. Each mode is different, and requires a different pattern of supports and driving location.