I agree on most accounts, but BE-256 certainly does attenuate like a Belgian high gravity yeast, minus the finicky nature of 3787 which sometimes stalls near 1.020-ish.
It does indeed attenuate like there's no tomorrow, leaving no maltotriose at all, which means only dextrins contribute to the beer's body and allowing you to create beers with the starting gravity of a small black hole that still come out very drinkable. However, I would hesitate to call this typical for a Belgian high gravity yeast. The attenuation of, say, WLP540 (originally taken from the Rochefort trappist brewery strain) is far more moderate, which creates a beer far closer to what I know as the genuine Belgian profile. I do know that in the US most Belgian styles are currently being presented in American presentations of the style, so BE-256 may be more appropriate there, but as far as the "native" Belgian beers are concerned, such a high attenuation is not typical. It is convenient for the brewer, of course, since other Belgian yeasts can take far too much time to ferment the beer out all the way, leading to overcarbonated bottle conditioned beers, but with BE-256, done is done. So from a commercial brewing standpoint the fermentation kinetics of BE-256 are preferable and give the brewer much greater control over the the final product, creating a far more stable beer. My biggest gripe with BE-256, other than the need to ferment unusually cool to prevent a banana explosion, is that it's POF negative which, in my opinion, immediately rules it out as an Abbey style yeast.
Wyeast 3787 / WLP530 (originated at Westmalle) is more in line with the traditional attenuation of Belgian strains, and is properly POF+. And yes, most "real" Belgian strains can be finicky, especially when it comes to completing the fermentation. How do those darn monks do that to get it right every time?
I have used BE-256 to ferment several VERY high gravity imperial stouts which finished from 17 to almost 20% ABV and had little-to-no yeast character. BE-256 regularly takes my 1.100 beers to 1.016 or lower which can be problematic because some beers like stouts can end up too dry.
BE-256 is indeed a great yeast for that sort of style! And that's exactly it: I'm not saying it's a bad yeast, because it isn't! It's just not an Abbey yeast. But then, in the past Fermentis positioned S-33 as a yeast suitable for Belgian styles, while it's really EDME yeast. Go figure.
As for your impy stouts ending up too dry, the answer would be to mash at a higher temperature and get some dextrins into the wort (or add some dextrin malt) since with BE-256 dextrins are your only option as far as residuals are concerned.