esp 101 with a variation

Hello everyone, I present my problem for which I ask your advice.
I have built a 101 by Rod and find it a good design, especially simple and good sounding with R 8 speakers.
I am now using Yankee audio planars which measured at rest give me R2.7.
I thought I'd use the vas part of Rod's 101 to drive three pairs of Hitachi 2sk135 and 2sj50, rebuilding the pcb to accommodate the six TO3s. I have 70V on the Hitachi power amps and 58V regulated on the vas to the MJE350 mosfet driver.
I set the bias to 200ma per supply branch.
The sound of the Yankee is spectacular, never heard such beautiful, precise and impressive midrange. After listening for a long time it seemed to me that the bass was a bit behind the music scene, a bit attenuated. I have two 800VA toroidal tubes and 90,000 micro caps per channel. I tried biasing the mosfets to 300ma per branch, but the bass went back even further.
I did some measurements, on a resistive load of 8 I have at clipping 37V about 171W
on a load of 4 35V about 306W. It seems to me that by raising the bias of the mosfets they start to raise the temperature and consequently increase the rds on. This could cause the MJE350 driver to fail and no longer be sufficient to drive the six mosfets. If this is the case, I have two options. First option is to leave the mosfets at 70V, which becomes 68V at work, and increase the vas and driver to 75V regulated.
Second option is to replace the MJE350 with a more powerful one.
Do you have any suggestions or have you already gone through this experience.
Thanks to all those who want to leave a contribution.
GIUBOS