[this has been tackled elsewhere, but not answered!]

Plug a drop light into GFCI - the light lights up!
Plug drop light into house AC - the light lights up.

Plug an extension cord into GFCI and test with
polarity tester - all is well with gnd, hot and neutral.

Plug (same) drop light into above extension cord and
the GFCI "blows" (i.e. cuts the circuit breaker).

The resistance of the extension cord only looks like a bigger drop light
bulb. So resistance is not the problem.

Plug a "cheater" into extension cord (above) and plug drop light into
the cheater (i.e. ground wire is not connected) making sure that there
are no possible ground contact points - GFCI "blows" [see above] !
So the grounding is not a problem.

When the GFCI compares the AC current in with the AV current out it is
obviously comparing similar phases (a.k.a. cycles). Clearly any phase
shift due to capacitance in the extension cord COULD cause what we see.

- Rick