As I think Sluckey is pointing out, a grid stopper would fix the problem.
If it were my amp, I'd add a 68-100kΩ grid stopper right at the 6SQ7 grid, between it and the junction of the 47kΩ resistors at the input jacks.
Yeah, I know it will reduce treble somewhat. But it will also reduce RF by a lot more, which seems like what you need.
Alternative: Is the tube loose in its socket? The metal shell is probably grounded via pin 1 (otherwise unused, many 6SQ7's and other metal-bodied tubes connect the shell to this pin). If the tube is loose in the socket, the shell may not be acting as a shield the way it otherwise would.
... I didn't see a reason for which a ferrite bead + an in-series resistor can't be applied ??
You can use both if you want. Like having a belt
and suspenders.
My be the presence of the resistor and ferrite bead at the same time push the frequency cut too high as to stop a low radio frequency, it stops only high radio frequencies
Both do the same thing. Because the ferrite bead is the same as a small inductance, there would be no impedance presented at d.c., and so no reduction of d.c. voltage. A resistor, in general, could cause a d.c. voltage drop. However, in the kind of tube circuits we're talking about, this is not a problem and there's no d.c. voltage division (because of the very high input impedance of the tube grid).
A ferrite bead would be technically superior, but you could use a grid stopper resistor and get just as good results. I've never bought a ferrite bead, so I have no idea if there is a price difference that would influence the decision.