Are you saying that transformers for high voltage transmission use Y configuration for both the secondary step-up driving the line and primary of step-down being driven? Unlike Y secondary and Delta primary for most low-voltage applications? Because it needs to drain off high voltage low current which is developed, to avoid breakdown of insulation?
That's my understanding, yes.
Seems to me a resistor would either allow the current anyway, or limit current but allow excessive voltage. So disconnect would be the only way.
Typical transmission lines run AC currents per-phase-conductor in the 750A ballpark during normal operation and the largest ones are rated a bit over 4,000A. The DC drift to hysterical voltages of elevated wires with no DC path to ground due to the atmospheric vertical field is driven by leakage from the upper-air/ground capacitor, and the bulk of the charging of that, for the entire world, comes from electrical storms - at about an amp each and about 1,000 active at any time. (A tad more comes from solar winds and cosmic radiation.) A transmission line intercepts a minuscule fraction of that current - and one of them could handle most of it for the entire planet. The "charged by the atmosphere" effect is in the hours timescale.
A ground fault on one conductor would drive the other two to a somewhat higher voltage (sqrt(3), about 1.73, times normal) and unbalanced loads among the phases produce unequal voltage drops in the three conductors and drag the voltage distribution off-center - which is what bypassing a resistor with a big capacitor is about: The ground fault and drag-back-toward-centered currents are AC, so the capacitors can handle most of that. Arcs are intermittent, though, so the capacitance of the wires to ground and the bypass capacitor, if present, would acquire some DC charge - a fraction of the operating voltage - from an arc / ground fault, and the resistor would drain that. This would typically produce far more load on the resistor than the atmospheric currents.
Lightning strikes, lightning-induced surges, and other very high voltage events are handled by the lightning protection gadgetry. Though magnetic storms might induce, say, five volts per mile, transmission lines typically are under a thousand miles long and run at 138 to 765 kV (and occasionally more). So the induced voltage is at most a single-digit percentage of their operating voltage, far less than the boost on two of the phases from a ground fault on the third.
The problem is that the mag storm voltage persists for a long time - many cycles of the AC. So if nothing is done to limit it, the DC-component of currents in the transformer coils builds up, the core saturates, and then the current driven by the line power voltage goes sky-high and burns out the transformer.
Still addictively interesting.
. 
