my thoughts on these cit- 12 amps is upgrade the power supply caps to 10000 uf 63v ( i used elnas ) the pots on these are very important to stay in tolerance , they actual value is somewhere between 90- 100 ohms believe it or not , replace with 500 ohm pots and recommend to adjust to .23 d.c volts after warming up amp!! Also the differential tranny is replaceable, my idea was finding two identical hfe ( doesnt matter what # you come up with as long as they are matched hfe value ) i used 2n2222 to-92 style) and i superglued those together kinda looks like a spyder still add .. those trannys with heatsinks on the pre-side are usually audiobly ok to use! same without drivers originals trannys are better to have! Q704, 704 replace with 512-kcs2690ays- Q701, 702 i used 2n5087- use quality caps and micro film resistors throughout too !! added 36k 2watt bleed resistors on each half of the powersupply, helps with the turnoff thump these a are famous for! eliminate inputs caps (220uf) - more bass without these filter caps! gold rcas and gold 5-way binding posts and better ac cable (mine was 14 gage with ground wire) check for any serious warmth (heat) while at idle that means bias is up to high- adjust pots lower!
If anything connected to the input has a parallel resistance to ground, it directly affects the first transistor base current so the differential stage is really in disorder leading to DC offsets and distortion.
This would happen if a preamp is DC coupled output, or AC coupled with a not too high resistor bleeding the preamp output cap.
Or, as
dshoaf stated, a Citation 11 is connected which shorts its outputs.
Apart from that, the input cap is high capacitance enough to avoid any impact on audible bass level.
So myself, I do NOT see any technical reason to get rid of the input cap.
Now the pots must be 4k7 not 500, if 500 you will indeed need to tweek down to 100 or so to sink these Voltage amp transistor 3 milliamps or so, and then indeed the pots setting tolerance is important because the q707 transistor does nothing (acts like open circuit).
If the pot is 500 or lower the base voltage for the transistor will not be met, so amplifying of temp diode voltage change because of temperature differences by the transistor is not there.
So no proper bias temp compensation although it still will compensate more or less, although not as per designers intentions.
Now, the designers intentions would be more or less obsolete anyway, using other type of output transistors having possible other VBE and junction to case temperature behavior...
these are "assumptions" again, I might be wrong so any comments are welcome.