Irrespective of a "house sound" and what may float one over the other's boat between different brands of amplifiers, there are ways to more effectively harness the potential of a given amp with how it's loaded. Looking into a passive cross-over of a speaker, not least a complex one and delivering the power over a full range spectrum under such circumstances is way more challenging and therefore more compromising wrt. performance envelope vs. seeing a pure load directly into a driver from a dedicated amp channel and over a limited frequency span. The amps functioning in the latter scenario will see their power and quality much better and easier utilized, whereas with the former the amp would ideally need to be close to impervious to load (which is saying a lot in the face of a difficult, full-range ditto) for any hopes of it to be a comparable scenario with the latter - and that is disregarding the sonic influence of the passive cross-over itself, one might add.
I'd wager a vital aspect of why a, say, $35k amp sounds better over a pair of speakers than its cheaper $20k sibling might very well come down to the fact that the more expensive amp is less affected by load and therefore has more power headroom/is more at ease operating, with all that entails. Indeed, passively configured and inefficient speakers with complex XO's call for the need of amplifiers with massive power supplies and overall sturdy build to come to life, whereas conversely in an active scenario much less is needed of the amps to still be fully up to the task, not least power-wise - again, with all that entails.
Configuring speakers actively myself I see no reason to strive for 10's of thousands $$ amplifiers when I can get by with much less. Moreover, this scenario (i.e.: active) lends the opportunity of a differentiated amp approach, where fewer quality wattages can be used in the mids to upper frequencies (even less watts the higher the speaker efficiency), and more brute force can be used in the lower regions where it's more readily required, in addition to the typical need for more damping factor here and proper driver control (this could be done passively as well with a bi-amping approach, but without the same opportunities of active to blend different amps more smoothly). Surely, when developing amplifiers it would seem that the marriage of the more ultimate in sound quality with gobs of power doesn't always go hand-in-hand, and thus differentiation of amp usage via an active setup can make even more sense.
Whether a $40k amplifier is worth the investment is up to each to decide. If it makes sense to someone to throw that amount of dough after it in the system context it's supposed to be used, go for it. It's not that I can't see the reasons why expensive amps are expensive, but coming down to it it's only saying so much: that's they're very expensive, and it figures why. I certainly wouldn't automatically assume they're the better for it compared to offerings much cheaper, depending on the context they're to be implemented.
