Why do tube amps often subjectively sound more powerful than SS ?

It’s because they’re old school sound, from the days when tubes were all there was.  Some of us were born in the days when SS didn’t sound so good. SS has gotten much better. But then again, so have tubes. Today’s tube amps aren’t your Great-Grandfather’s tube amps. 

Actually you don't need a degree to design tube amps.

@dynamiclinearity That is certainly true. But it helps, especially if the amp in question has a feedback loop. Designing a good feedback loop is a bit less trivial if you want to get it right.

My point was there's been a resurgence of SETs since the early 1990s, most of which don't use feedback. They tend to have a high output impedance and so speakers that expect the amp to behave like a Voltage source (which is most speakers made) don't work right with them. Hence along with SETs we have more horn and open baffle designs meant to work with low power tube amps with high output impedance.

If Stereophile was measuring such an amp it wouldn't measure well by their standards but might sound pretty decent.

Some tube amps I would expect to have a higher damping factor are those made by Roger Modjesky.

Futterman made some OTLs with quite high feedback and had damping factors as high as 40:1. You can put more feedback on an OTL since you don't have the poles created by the output transformer; you can have a greater phase margin. Khron Hite made some laboratory amps (UA-101) in the early 1960s that had as much as 80dB of feedback (and hence had very sophisticated feedback design) which had a damping factor of 100:1.

I had demonstrated to me by my friend Murray Zeligman and David Berning that full loop negative feed back was deleterious to linearity(a Berning amp with variable feedback and as the feedback was reduced to zero what I heard sounded better to me).

@dynamiclinearity  If the feedback isn't set up right you will get exactly this sort of result. One of the problems is the feedback is sent to the cathode of the input tube. The tube isn't linear of course, so the feedback signal is distorted. This causes additional higher ordered harmonics. Norman Crowhurst wrote about this problem nearly 70 years ago but its rare that anyone has done anything about it. 

The solution to that problem of course (one we've used for decades) is to apply the feedback to the grid of the tube rather than the cathode, in a manner similar to how its done with opamps. The feedback is thus mixed with the audio in a resistive divider network which is far more linear than any tube or transistor. This reduces the harmonic and intermodulation distortion generation you otherwise get. Of course you have to sort out how to get the phase right but if an output transformer is involved its easy.