I have NEVER had a piece of radio communications equipment malfunction or self destruct due to coax length. Hmmmmm, strange.
Not strange, just lucky.
The difference between a RF amplifier and a RF oscillator is how you control regenerative (positive) feedback. Any amplifier that has a poorly-designed or botched input circuit is a potential oscillator. ALL amplifying devices will feed some of their output back to their own input, INTERNALLY. Add to this the tendency of RF to just magically travel across empty space inside the amplifier, and this just increases the overall unwanted feedback, over and above the transistor's built-in feedback.
The input circuit serves, among other things, to shunt unwanted regenerative feedback to ground, away from the transistor's input terminal. The lower the impedance at this spot the better, up to a point. The lower it is, the more feedback current it carries safely to ground, away from the transistor's inputs.
A coax jumper makes a dandy tuned circuit. Placing a tuned circuit across the input of an unstable amplifier is a good way to construct an oscillator.
Coax length should make no difference whatever to an amplifier with a perfect resistive 50 ohm input. None. Zero, Nada.
The farther an amplifier's input circuit gets above or below that one-to-one SWR impedance, the more difference coax length will make. An amplifier with a five-to-one SWR measured at its input while keyed will almost certainly change behavior as the jumper length is altered.
73