Output transformers for high power PA amplifiers have been in use for many years and their response is entirely adequate. A transformer provides a very simple way to select taps to adjust to the particular load impedance and provides the benefit of complete galvanic isolation between the amplifier stages and the load. Thusa transformer will be the simple solution.
Given that the commercially available ten watt amplifiers must be working well enough, the calculations showing a requirement of 500 watts must have used some incorrect assumptions, it appears. A ten watt amplifier is far more believable than a 500 watt unit, which could not run off a set of AA batteries for any useful length of time. Thus we see that what is most likely is that the ten watts, 63 volts, and 370 ohms, do not all happen at the same time, but rather that at some point the pipe to earth impedance is 370 ohms, and that under some conditions the applied voltage may rise to 63 volts, and that at none of those conditions is the delivered power ten watts.

 

@MisterBill2

Each parameter casts a limit of some sort. I learned that the hard way. The power rails. The max current. The load.

So, I bumped down a built knowing that I built the Leach amp. It has +-50VDC rails. It was spec'ed for 72 W which is even wrong. I^2*R or 9*8 = 72 W; I=3. I, primarily limits power.

A bi-polal 50V rail should be able to swing +-42V p-p into 320 ohms.

It's not a 10W amp and it's not a 500 W amp.

The power supply was fine for electrostatic speakers.

At 30 V RMS into 8 ohms, you can get around 110W

If the load changes to say 2 ohms or to 1000 or the available current changes, then all bets are off.

There's a lot of RMS, peak to peak, Average, load values thrown around. The numbers thrown around are 63 V RMS, 10W, 25 mA and 320 ohms,

 

A line to voice coil amp might be a way to head. https://help.chimemaster.com/p/70V_Line_Speaker_Wiring

See also: https://www.electronics-tutorials.ws/transformer/audio-transformer.html

This http://www.hammondmfg.com/pdf/5c007.pdf is always a useful read.

 

I was thinking about a transformer that matched the amplifier output to a bunch of different impedance values. The match does not need to be exact to be adequate. Winding an audio transformer with a bunch of taps is not that hard a task. And at a ten watt power level it would be an easy to handle size.