I've built two Princeton Reverbs and a Super Reverb, so I've got a couple of thoughts. Keep the leads going to and from the Intensity pot as far away from everything else as possible. I run them over the board and have 'em drop straight down. Shielded leads to and from the Reverb pot reduce noise a lot. That's a sensitive, low voltage part of the circuit. An audio taper 25K Mid pot gives you a mild "raw" boost if you want it. Reducing the coupling caps between the PI and the power tubes may be good for reducing bass, depending on your OT, speaker & cab. Lowering the value of the cathode bypass cap on the reverb driver (12AT7) helps keep the reverb from getting boomy. Add 470 ohm (minimum) screen grid resistors. 1.5K grid stoppers on the power tubes also.
Adjustable bias is a really good idea in this circuit, especially with the bias vary tremolo. There's a sweet spot where the bias is cool enough for the tremolo to work really well but hot enough to sound good. I've got a nice circuit and layout for the bias board if you want it, but you'd need to send me your email address.
About the grounding, it's virtually impossible to use the Hoffman approach with the JJ 40-20-20-20 cap can. I did it by adding filter caps for the preamp and not using all of the cap can.
Some guys like to reduce the 3.3meg reverb mixing resistor to 2.2meg or even lower. It makes the "dry" signal stronger relative to the signal from the reverb return.
If you look at the schematic, there's an "unused" node on the power rail between the two 18K resistors. Again, some like to move the plate supply for the PI to this higher voltage node with the goal of increasing clean headroom for the concertina part of the PI. HBP and I have discussed this "Stokes Mod" at length, but I have not proven through experiments that it actually works as advertised.
That's all I can think of off the top of my head.
Cheers,
Chip