A recent look at a Harbeth XO by GR Research found about 30 components in the XO! Unfortunately, although this can produce a nice flat frequency response, which will satisfy potential buyers but more importantly be favourably mentioned by the reviewers, will reduce dynamics and micro-detail. Low level nuance also suffers.
This kind of stuff is often claimed in audiophile circles. But then so is lots of "stuff."
I'd like to see this actually demonstrated, and not just by some audiophile's say-so.
If we are talking regular old anecdotal subjective listening impressions, Thiel's complex crossovers used to get grief from the "don't put too much stuff in the crossover region because blah, blah, blah" crowd. Yet every Thiel I've had has been world class in it's price range in terms of detail, soundstaging, timbre and dynamics.
Personally I don't rely on audiophile shibboleths - the result is what matters.
And as for the parts quality in the Harbeths (which I have a fair amount of experience listening to), the results are what count. Engineers will point out that good or clever engineering isn't just "what you can do if handed the best possible parts and budget" but rather being able to achieve excellent performance without having to rely on the most expensive parts. Which is what Harbeth seems to have done.
Dan Richie in his videos does some interesting and entertaining and educational stuff. But I am not inclined to ignore that he is a salesman with stuff to sell, and arguments for his own products and services, as much as anything else.