Speaker power compression IS another important topic and one that is overlooked by the 85db/watt speaker crowd. The truth there is that compression sets in far earlier than people would expect and this has the effect of blunting live dynamic swings. It also means the speaker can quickly become a bottomless pit for power as doubling of power no longer gives 3db increase in volume. There was a demo on video (maybe youtube but I don't remember now) of Harbeth speakers and CH Precision amps. The amps were showing some ridiculously high power values (over 600 watts if I remember correctly) and the guys were still able to stand and listen and even talk during the playback! Now, those speakers are somewhat low sensitivity but with that much power going in them it should have been quite impossible to get close to them or talk. This tells me that this speaker was simply not getting louder anymore regardless of the power you put in and that it would likely blunt realistic dynamics at anything above background listening levels. Sadly, the majority of conventional box speakers are nearly like the Harbeth and will never achieve life-like dynamic contrasts unless the average level is around 70db but then you are usually losing out the realistic contrasts...at the low volume end of the spectrum because insensitive speakers drop that out as well. A typical 85-88db box speaker is ideal for average levels in the upper mid 80s db wise with a small dynamic range...in other words they are ideal for compressed pop/rock music. Once a box speaker gets above 90db/watt they start to get kind of interesting for less compressed music as long as they have their driver/cabinet colorations under control. Once they get above 94-95db they start to get interesting for wide dynamic range music at something approaching realistic levels, assuming that they have the whole frequency range under control (many horns do not).
If you put high powered SS amps on a high sensitivity speaker of high quality it will not sound more dynamic or alive than a good SET or push/pull triode amp. It will also not sound more powerful but you will probably end up having to turn it up louder for it to sound less flat and boring.
I have found though that electrostatic speakers and to a lesser extent ribbons are somewhat different in how they relate to power. They don't seem to suffer from power compression, perhaps because they have such huge surface areas acting effectively as voice coils and as a result don't heat up really. Also, as a line source, they have much less drop off with distance in a room so the in-room sensitivity is higher than what one might expect. What this means for electrostats is that they go down down down in level very nicely without dropping out the music and when the average level is set at something reasonable BECAUSE you can hear down in level so low this means that you can achieve a realistic scaling of the signal up until the excursion limits (with the bass usually) kick in. This means you can do quite realistic reproduction of classical chamber music and small acoustic jazz ensembles at realistic levels (as long as there is not too much bass) and can get realistic scaling of big orchestral works...just not at live levels...perhaps 6-10db below live. Ribbons are a bit more limited on the low volume end, needing a bit more juice to come alive but they might be a bit more capable at the high volume end, although the bass panels will bottom out in about the same way as a big electrostatic panel will. I have found that there needs to be a few db higher average listening level with ribbons for high dynamic range music, which means it might not track the big orchestral dynamic envelope as physical compression due to excursion limits sets in.
My friends and I have heard this as well even on speakers like electrostats and ribbons where we have been quite successful running them with push/pull triode and SETS. I was running Acoustats for years on KR Audio SETs and it was very live sounding...much more so than my friend who had a pair with a large ASR Emitter II Exclusive or Musical Fidelity KW750 (not many amps with more clean power than that one) or a number of Krells, the Lamm M1.1s, Sphinx Project 14s, SimAudio Moon W5 etc. Shouldn't work well according to the power firsters but it did and so well that some of those friends followed suit. Now, I have horns (Odeon "La Boheme") driven by SET but I have tried push/pull tubes and transistor output hybrids on them...SET rules for dynamic expression. Apogees can work fantastically with SET and PP triode amps. Much more life there.
Basically, all my local friends have dumped their high powered SS amps for SET or PP triode. They had amps from the likes of McIntosh (MC 501), Karan (big 200 watt integrated), Krell (various), Sphinx (Project 14), Lamm, Musical Fidelity, Sim Audio, ASR etc. They all believed the power hype...not anymore. All are much happier with the sound they get from tubes (SET in particular). Speakers include, Odeons (three different models), Apogees (Grands, STudio Grands, Centaur Major), Acoustat (1+1, Spectra 2200, 4400), Thiel CS3.7, Stax ELS F-81, Cabasse Baltic Evolution (above 80Hz), Reference 3a (3 models). Note, the only low sensitivity speakers here are ribbons and Stats. The Thiels are 91 db and work great with SET and PP triode (VAC 30/30 for example) the Cabasse are about 89 db but an easy load. Ref 3as are around 92db and easy loads and Odeons range from 93db (Rigolettos) and No. 32s (95db) to 98db (La Bohemes).
So, one has to ask oneself, what is the big difference here in power amp design?? Turns out that what you usually have to do to get ultra high power with ultra low THD is a sound and dynamics killer.
