Train -- Here's something that will help you out as well. The X-101ST did run the output tubes quite hard -- particularly the early version that has ganged rather than concentric tone controls. However the output stage of the X-101ST is nearly identical to those in the TA-600, one of which I happened to do some extensive work on for a fellow AKer. You can find that thread here:
http://www.audiokarma.org/forums/index.php?threads/ta-600-service-bulletin-early-version.676168/
With the audio circuits of the TA-600 so heavily lifted from the X-101ST (except around the phase inverter/driver circuit for those versions using a 7199 in that position), you may fine post #47 (circuit) and post # 34 (installation) most interesting in this thread, where I installed an IBAM for this model, because it too was red plating the output tubes with nothing wrong other than the stock design requiring a very tight match on the quiescent current of each output tube. If one doesn't conduct it's full share of the load, it leaves the remainder to be spread among the other tubes, which then causes them to operate in an over-dissipated condition. As the match becomes further off from ideal, the red plating becomes worse in the remaining tubes.
The heavy current draw is required because like the X-101ST, the TA-600 uses output stage cathode current to light the phono preamp tubes heaters. With a typical plate-to-cathode voltage of nearly 400 volts in these models, and a total cathode current requirement of about 140 mA, it pushes the tubes right to their Design Center rating for dissipation: 140 mA is 35 mA cathode current per tube. Subtracting out 5 mA per tube for screen current, it means that each plate is passing 30 mA. With a 400 volt differential between the cathode and plate, this produces 12 watts of plate dissipation -- which is fine for these tubes, but because they all need to operate at this level, their matching is of utmost importance. In today's audio environment, such tight matching is usually not possible, so some sort of IBAM is required for safe operation today.
The IBAM circuit I developed allows each tube to be adjusted individually of course. It works very well, but due to the biasing scheme used in general, the controls are somewhat more interactive than when an IBAM is applied to a fixed bias design -- particularly when used with a SS power supply.
Installing a negative bias supply certainly works as well, but of course requires more modification and an additional transformer which this approach does not require. You will also likely need to look at the resulting screen grid voltage with such a modification as well, to make sure that doesn't get exceeded under dynamic conditions. In any event, the approach I used in the TA-600 provides a very workable method of installing an IBAM to these units -- as will adding the negative bias supply as well -- so the current load can be balanced equally between the four tubes.
Dave