With a more saggy supply that thing would make for a nice guitar practice amp. The low order monotonic harmonic distortion profile sounds more pleasing than pp amps, which inherently filter out the even harmonics, making them more edgy and biting (they're better for cutting through the din in live gigs though).
As for using it in an audio system? I'll second the other's opinions that it's a bad idea. If you want the closest thing to "straight wire with gain" performance, go straight to a purify or hypex based class D amp.
Specifically for that amp, per the measurements in the ASR thread, the supply is clearly not up to snuff, as psrr is critical with these single ended class A SETs and that one exhibits way too much hum, and the output transformers are also sub-par and limiting bandwidth. It's cheap chi-fi, so no surprises. Dude also should have included some square waves in the measurements, they reveal a lot.
The measurements do reveal some aspects of the amps performance in reductive terms. Of note is the low order monotonic distortion, which, if only occurring in greater quantities on transient peaks, you'll never notice as distortion, but rather as increased loudness. This compressor-like behavior is the root of the myth that "tube watts are more powerful." It's really just more ear-friendly clipping behavior. And as with any compression, it can bring up low level details. I think all that would account for the positive subjective impressions from the reviewer.
Back to your specific questions, don't waste time on a tubed pre. Those can be as squeaky clean (or as dirty) as desired, feeding a fixed high impedance load, but the proverbial "tube sound" results from a high source impedance amp interacting with transducers of varying load at different frequencies, with all the power limits/clipping behavior and distortion that comes with it. Unless your speakers are sensitive enough to meet dynamic range goals at 1.75 watts maximum, it's a no-go. Your Polks aren't such speakers.
