In both cases, they do not look like grid stoppers to me. The value, at 68K, is much higher than necessary. In the "...question3" circuit I seeone input to two triodes, one of which has a much smaller bias resistance than the other. Without the plate loads it's hard to tell, but quite possibly the bias voltage is much lower on that triode, allowing it to be driven into some grid current on signal peaks. The resistors would prevent that from happening identically to both triodes.
In the "...question2" fragment there are two inputs to one triode. The resistances would prevent interaction between the two inputs, in particular the risk of loading them with the other pickup's impedance. This is a fairly standard approach to multiple inputs which you want to add together.
The 12AX7 is not a very high transconductance tube, and thus grid stoppers are less likely to be needed. I use them everywhere as cheap insurance, but it's the high transconductance tubes that really need them.
