In the past several months I have done hours of listening and measuring of my 3.5 with passive and active digital xo. That Thiel originally also explored the triamping active xo makes good sense. The passive xo system still has these problems:
Each speaker has its own min phase characteristic. Minimum phase still has phase change depending on frequency range although "minimal". Only linear phase has a constant phase change independent of frequency range. Overlapping frequency range of multi-speaker systems still introduce phase changes especially with the 1st order xo that allows large overlapping ranges.
The variability of components is not negligible and when measuring each component the factory specs are only an approximation and fine tuning is essential. Taking good and reliable measures is an advanced skill but also frustrating as the components can change characteristics based on temp, load, etc. Then the issue of what differences are audible and not determined by subjective bias requires double blind comparisons. A passive xo system modification is a fair amount of work assuming that needed components are really available. My conclusion so far is that active digital xo system have the best chance to approach the ideal time phase coherent system with linear phase, time aligned, component eq, and even individual room adjusted set up. As it is stated that the speaker and room variables account for the largest part of the reproduced sound quality, it only makes sense to focus any innovations in mastering the active digital xo systems.