1. The 470KΩ resistors are wired from the input signal connector to ground and establish the input impedance for each channel of the Stereo 70. I believe they might also function as grid leak resistors for the input to the gain stage (first section) of the 7199. From the Aiken Amplification glossary:
"Grid leak resistor - a very large resistor from the grid of a tube to ground, which is used to generate the bias voltage for the tube. See "grid leak biasing" for an explanation of how this works. This term is sometimes incorrectly used when referring to the grid-to-ground resistor in a cathode biased configuration, which is used to provide a DC ground reference for the grid circuit."
2. "So any voltage change with the larger caps can be offset by the bias pots." I would not expect any significant change in voltage due to larger value bias capacitors. Difficulties in adjusting the bias pots sometimes happen when a silicon diode is substituted for the original selenium stack possibly in conjunction with the AC mains being too high or too low.
3. Biasing the 7199s occurs with the associated resistors. As long as the resistor values are in-tolerance, and the power supply voltages are correct, and the 7199s are good samples, they will bias up properly. The 82pF does not block DC to the phase inverter grid, because it is not in series with the connection. There should be no appreciable DC current flowing into the grid of the phase inverter (at pin 9), as the input impedance of the grid is extremely high. I believe the 82pF is there to siphon off any ultrasonic oscillations, or something to that effect.
4. Pins 2 and 9 are connected together to couple the audio signal from the first section of the tube (gain) to the second section (phase splitter/inverter). I assume there is no coupling capacitor because the DC voltage at pin 2 helps to bias the phase inverter triode. The voltages on the two plates are different because they have different source voltages from the power supply and different value plate resistors.
5. Testing the choke out-of-circuit should result in about 1.75 Henries and 62 ohms of DC resistance. Actually, looking at the schematic, you should also be able to take these measurements with the rectifier tube removed, assuming the circuit is in the original configuration.
6. For the 270KΩ resistors, notice the asterisk. This means they need to be a 1% matched pair. I believe this is so that the bias voltage (injected at eyelets 6 and 21) is evenly divided to each pair of EL34 tubes - there's only a single bias adjustment pot per channel.
7. The 1KΩ resistors at the input grids of the EL34s are "grid stopper" resistors, and they are not there to set the level of the signals going into the EL34s. In terms of the audio signal level, their value is not critical because 1KΩ is negligible compared with the very high Z-In of the EL34 input grids. The grid stoppers are used to prevent "parasitic oscillations," which gets kind of technical: http://www.aikenamps.com/index.php/grid-resistors-why-are-they-used
8. The return path is through the EL34s.
If I have anything wrong here, please speak up. Thanks.