Bingo @moonlighter @peteb , the soldering to the copper plate was gone, I've fixed it.Okay, so based on that info we're drawing about 96ma between the plate and screen grid, just as you calculated before... that 47V is more than double what we should be seeing (21-22V).
That cathode resistance is plenty good enough, so we'll stop worrying about that.
Pin 5 to ground should measure 220K. If that is reading open, we do have an issue because the tube has no grid leak. This could very well be the problem. Without a grid leak, electrons pool on the grid and increase the bias (in power tubes this is dangerously drastic). The 6V6 datasheet specifies a maximum of 500k, and Fender used and plenty safe 220k.
I measure as follow:
B cathode voltage 23V
Bias current 44mA.
Tomorrow I will play it and see how it sounds.