When a budget speaker is preferred to a high end one...

In my systems I have rarely if ever experienced a speaker at 1/3 the cost trumping another. However, in others' systems I have. As others have already mentioned much has to do with room placement, personal preference, room treatment application (or lack thereof) and system synergy. What's also key is the components' quality, or more specifically lack of quality. More expensive speakers in general will be more revealing, fleshing out weaknesses in gear so it may sound more fatiguing, may exhibit loose bass, etc. making the less expensive speaker preferred. Less expensive speakers commonly have "sins of omission", which may be a good thing depending on your system.  

@audition__audio wrote:

Yes other speakers sound anemic because I think you like all of the information being added by the Harbeth. [...]

Harbeth was really only an example of a relative minority of "hifi" speakers that reproduces named power region with a more proper richness and energy. As a said a less fulsome presentation here emphasizes upper range detail and the lower end (the "hifi" imprinting), but at the expense of a more natural, coherent and live sound - to my ears. 

Well it seems to me that Shaw has said that all amps sound the same through his speakers. To me this statement diminishes the obvious importance of other components and elevates the speaker in terms of significance. This also runs contrary to my experience. 

It's a radical statement by Mr. Shaw, indeed, and it also ties into his assessment on the importance of cables and "burn-in." He's much the engineering mind, but if Harbeth speakers in general are a relatively benign load as seen by the amp it would certainly diminish its importance (a tendency only exacerbated with active configuration sans passive XO's) - not that I'd go so far to say they all sound the same.

I do find it's freeing for a man in his position to not enter the wagon of "everything matters"-craze that can't identify more important "pillars" of core aspects in audio reproduction to focus on. 

I think Shaw is dead wrong about thin walled cabinets and energy dissipation. Sure the energy is removed but by the vibration of the cabinet which adds audible colorations. 

I would refer to poster @cd318's informative reply just above.

His crossover designs run contrary to my belief that first order crossovers are the best compromise and that complexity in this arena is a really poor idea.

My quip: if it works it works. 

I think it is also interesting that much of what Shaw proposes also reduces the cost of his product significantly which, considering the sum of its parts, I find very expensive. 

There appears to be a clear rationale at work in the service of natural sound found in Mr. Shaw's rather analogue and straight forward ways of identifying what a voice actually sounds like and how to come about reproducing it fairly authentically. Few speaker manufacturers, as I see it, are really interested in tonality and dynamics and how to take the necessary steps to get these vital aspects in sound reproduction (in a closer direction of) right. The current, or even latest decades of paradigm seems to have technology dictate design choices from a theory-laden approach (that isn't without merits) much more than designing around a pragmatic outset that would have the close attention to naturally produced sounds pave way for the designs themselves.

If Mr. Shaw can get away with a more natural sound (whoever may agree on this) at reduced material costs, he isn't at the short end of the stick I'd say.

I think that the crux of what I am trying to say about the Harbeth sound is that the way the speakers are designed excludes the possibility of neutrality and "natural" sound.

I dont dislike Harbeth speakers I just object to the adjectives often used to describe their sound. With respect, I think that this extra sauce is what Harbeth owners like but I dont think that it is as true to the source as other speaker designs.

I was listening to two different floorstanders, I forgot which models. One retailed under $3000, the other $1000,  both used. The former had detail, frequency extension, etc., but sounded buttoned up, almost inert. The latter didn't have the precision, but it sounded more generous, more engaging. I preferred the latter.