More power for moderate listening levels?

 I added a T+A power supply that adds 1800 watts.

Power is V x A.
Increasing 4Ω power from 500 to 1800 requires changing the voltage from 45 to 85. Current would increase from 11 to 22A. It very unlikely the output devices were that over spec'd and the heat almost certainly wasn't.

FanBoy prattle.

Tekton makes some great speakers

IF one likes that sort of thing.

In the Monitor review, Stereophile determined "My estimate of the Impact's sensitivity was 87.5dB(B)/2.83V/m, considerably lower than the specified 94dB." https://www.stereophile.com/content/tekton-design-impact-monitor-loudspeaker-measurements

Truth in advertising: Miller owns Tektons.

I have found that a high powered amp, even at low listening levels, provides a sense of dynamics that a low powered amp doesn't have

And I'm fairly certain that the amps are likely sonically chalk and cheese.

As more power is fed to the voice coil, it heats up and that heating increases resistance which reduces the amount of power that can be delivered to the driver. The less efficient the speaker, the greater the compression.

Horsefeathers. Very efficient speakers have wimpy motors that may not control the cone as well as a less efficient speaker. The voice coil has to get to ≈90°C for a 1db reduction. While voice coil heating effect is measurable, it is difficult to detect. Our ears begin self-preservation clamping at elevated levels, which some may interpret as voice coil heating compression.

see http://www.ielogical.com/Audio/WinterBlues.php#TheFly for a bit on voice coil heating.

@brianbiehs:
Install an app like AudioTool on your phone and get some real measurements as to the levels you are reaching.

25wpc at 2.6m [8.5ft] max out @ 95db PEAK on 86dB/W/m speakers. At 5m [16.5ft] max level is 89db. Add room noise from a bunch of people and the apparent level will be somewhat lower.

Be aware that changing amplifiers to support Disco may severely compromise audio quality at less taxed listening levels.

More speakers are blown by party levels than almost all other causes combined.

There is no free lunch!

I'm hoping the higher power amp (15W Class A before switching) shares some of the sonic character of the current amp.

Class AB amps don't 'switch' per se. The slide into Class B. Depending on the design and the load, the transition can be smooth and less so.

Getting AB bias currents exactly right as the amp plays is an engineering challenge. It almost always lags the amp's thermal state. Manual bias amps are frequently mis-biased from the factory as the amp is set up with the covers off in an environmentally controlled lab. Close it up and stick in a rack and all bets are off.

see https://www.zenmod.in.rs/pass-labs-x150-5-checkingadjusting-offsets-iq-and-gain/ for a Pass 150.5 bias adjustment tutorial.

Like many other recent 'inventions', 15W Class A before switching is nothing more than ≈30mV bias across 10x 2Ω emitter resistors in ± amp halves.

Pass is a brilliant designer with an impeccable pedigree, so his designs are likely SoA.

That has got to be a lot more than a few mV bias.

The JC1 has 9 pairs of output transistors.
25W ≈ 22mV across 18x 1.3Ω emitter resistors