The problem is that the electronic circuit is, say energised and has a detection feed from the output, when it's the correct way round and it trips or is switched off, the detection circuit is dead, but the main circuit is still energised. All good.Do you have a mechanism in mind where it wouldn't detect an imbalance between the line and neutral lines which are AC ?
Unless it specifically calculated the phase angle (why do that ?) then it would see the same imbalance whichever way the device is inserted.
So ?
The detection circuit is normally energised 100% of the time. If you mean the trip circuit, when it's tripped, the fault will disappear. The only exception might be a N-E fault with a single pole device that doesn't switch the neutral, but who fits those ?
However if the wrong way round, when it trips, the detection circuit remains live, but the main circuit is disconnected! This can cause a reverse supply to the circuit. If it wasn't designed for this condition, it can damage the electronics.
Once damaged the rcd may not work - yet outwardly all you did was switch it off and back on again.
Only the manufacturer would know if keeping the detection circuit energised whilst the main control circuit is not causes a problem or not.
It isn't just RCDs BTW, RCBOs and especially combined AFDDs may be especially susceptible to this issue.
In reality if a manufacturer dictates their device is not suitable for bidirectional applications, they don't have to explain why; we need to follow their recommendations.