The bias cap is grounded twice? And there is a jumper going from the resistor hanging off the 6v6 sockets to ground? Both of those resistors are grounded, so that jumper seems redundant.
The solder lug that the 1 ohm resistors are soldered to are isolated from the chassis denoted by 'iso'. The ground jumper from the isolated resistor
is intended to pick up the bias cap on it's way to the B+2 ground node.
Does that make sense, or am I missing your point?
What I am aiming for here is a ground scheme that has only one ground to the chassis through isolating all other points.
Probably not necessary, evinced by numerous other builds, but this is a learning exercise as much as "I'd love to build a Harvard'.