I just want this to be simple enough that modeler makers don’t overthink it to the point of omitting it.
Oh, absolutely! And I'm sort of in the "anything is better than nothing" camp. Things should also offer the easiest and most transparent access from a user's POV.
Yet, I think there's at least some things that really need to be taken care off.
My main gripe would defenitely be to avoid anything that could end up in users doing tweaks to patches accidentally just because they've forgotten to deactivate certain global blocks. The Boss system defenitely lends to these things lurking around the corner all the time. I'm not exactly digital- or Boss-illiterate (in fact, I'd say it's kinda like the opposite), and yet I've been running into issues one or two times.
The next big issue (not as big, though) would be organizing global blocks. By now, after thinking about it yet some more, I'd pretty much strictly vote against any kind of "sub preset" system or whatsoever. The reasons being:
1) I would not like being limited by the overall number of global blocks - for instance, I'm actually running into the Boss limitation of 10 presets per global block only for the parametric EQ because my idea was to have one global EQ per each Tonex One preset I may ever use (fwiw, this approach is working extremely well), and with four slots already used up (two acoustic settings, two lead leveling settings), I've only got 6 left. Not good.
2) I fear one could end up with an enormous amount of these global block presets - and you'd really never know when it was safe to delete any of them. Unlike there was a kind of managing system, informing you about a) whether a global block preset is used, b) how many patches are using it and c) which patches they are. Again, I'm already running into issues with the GT's way of doing things as there's some global block presets where I simply don't know whether they're still in use in some banks. Now, I started documenting things ever since I was running into that, but it shouldn't be required doing such things.
Adding any kind of managment system however might be a lot more of a programming issue than tackling the entire thing in a different way straight from the start. And it might still not help the user much, simply because you might never want to delete any global blocks.
As a result of all of these, I still think a group-based approach would be the best thing to have. In case you don't assign a patch to any group, you wouldn't even see anything related to global blocks. Nothing to select, nothing to maintain, it'd be as if the function didn't even exist.
Also, in case the default patch copying routine would automatically remove the patch out of the group (unless you'd check a certain "keep patch in group" box), you'd never run into any issues of accidentally altering some precious settings because you're fooling around with that copied patch (which is exactly what happened to me on the GT).
It'd also allow you to use as many global blocks as you want (sure, there could be some kinda limit such as 1000 or whatever), they would never clog up any menus.
I'm sure there's several ways to skin this cat, I also wouldn't happen to know whether there's yet some better solutions than groups (let alone I wouldn't know how things would actually be realized using a group based approach, I'm neither a programmer nor UI designer), but the issues I've mentioned above are pretty real in my book and should better be adressed straight from the start.