Seems to me that the most fool-proof solution is the line tension sensor and a little bit of accumulative logic.

We had a discussion on the club group about this and the example used was a 1800lb tow-plane, 700lb glider being towed. Kiting event takes 2 seconds, ends up with the glider gaining 100ft. Starting speed 60mph (glider being a 1-26). After 2 seconds of glider kiting the glider speed is 75 mph, tow-plane speed is 40mph. Average tension on the rope is 900lb and the tow-plane experiences 1/2G deceleration along the longitudal axis. One could just set a longitudal axis g-meter and alert/cut the line at .5G, but that might trip with slack rope etc. Having a rope tension sensor calculate energy loss due to the pull on rope over the last .1, .2 ,.3 seconds would be easy. The hard part would be to decide what the limit should be. One might test this with boxing the wake, tows trough rotor etc. If the event starts fast, the rope angle doesn't really get much above normal (in reference to the tow-plane) since the strong pull on the rope easily lifts the tail. The energy loss for the tow-plane is the bigger problem.