That's pretty much my understanding also, that ratio deco is useful for planning deco from a significant deco dive, but that it doesn't have much relevance to the NDL or minimum deco (or in a few cases minimal deco) dives that we have been discussing.
One question still on the table is how well depth averaging works on dive profiles with a wide spread in the bottom depths.
For the 100' + 40' dive, as I noted above, I just treat it as a single level dive at 100' with a long hang time afterwards. As you note, a 100' + 80' dive can be treated pretty much as a 90' dive.
What happens though in a true multilevel dive, such as when one pops down a wall, then spends the entire dive slowly ascend back up. Is depth averaging applicable to planning such a dive?
I don't have enough experience with depth averaging of that sort of dive, to definitively say one way or the other. If I had to guess, from the few cases I've looked at in detail, I would say that it does work, but becomes unecessarily
conservative when the depths involved have a wide range.
Of course, failing by becoming too conservative is a good failure mode. OTOH, until I get a better handle on how depth averaging treats a variety of profiles, to use it would be much like doing a "trust me" dive.