AFAIK it works by lowering the signal/noise ratio to the point the channel is useless for it's intended purpose.

Something with extremely tight focus via aimable directional antenna or phased array, along with spread spectrum hopping would be extremely hard to jam, because you need to get enough energy into the channel to mess with it.

GPS jamming can be countered by a better inertial nav system, and even the tightbeam comms themselves can be used to locate approx positions with enough links.

Regarding combat capabilities, is there any fundamental reason for their relative lack of performance beyond it being unnecessary/risky with high-latency comms? From a quick scan of current costs, you could get >10 Avenger/Predator C drones (~15M ea) for a single F-35 (~200M) or F-22 (~150M).

I suspect (but am utterly unqualified to say) that such an encounter would go badly for the (human) fighter, especially if everyone has modern AMRAAM or similar BVR weapons ("... we only need to be lucky once").