[chrisdirkis]

For something to be a useful hypothesis, it must both match historical data and be good at predicting future data. Given that the signalling explanation is usually used post fact, and when used to predict is wrong as often as not (in my experience), it seems like a pretty useless hypothesis when used broadly.

Possible that there's a narrower hypothesis, or someone that uses it to accurately predict systems and behaviours. Would like to see it in that case.