I'd like to offer an analogy to VA that I think is illuminating, because there are similarities. That analogy is SLD (Supercooled Liquid Water Droplets).
Unlike VA, SLD has been linked to actual hull losses.
Similarly to VA, there are NO accepted criteria for SLD.
The only advice any OEM gives anyone regarding SLD is the same as that for VA a month ago - avoid at all costs.
There is no reliable means today for detecting SLD, or VA, onboard an aircraft. In certain circumstances visual detection of either may be possible, but is assured in neither case.
Our ability to reliable predict SLD or VA is, in a word, unreliable, in both cases normally for lack of accurate data about the atmosphere.
To date the situation with SLD constitutes a (barely) acceptable risk - although it's been an issue of contention between FAA and NTSB for years. Industry and the regulators have been trying to come up with workable means to handle SLD since Roselawn, in earnest, and have got not very far. We've largely "got away with SLD" because it tends to be a localized event.
So, regarding the comparisons:
IF we've lost a number of aircraft to a known but rare phenomenon (SLD) and havent managed to significantly mitigate the risk, what chance was there really of anyone being any more prepared for the VA issue of the form it took last month? And, realistically, what chance is there of being any more prepared a decade hence?
AND, suppose some unusual weather pattern happened to generate large areas of likely or possible SLD. Would that really leave authorities any choice but to close the airspace which represented a significant risk of SLD?
Oh, and before someone cites the existing icing regs: the environmental data they ultimately rest on is OLD and geographically concentrated, and may be unrepresentative of world-wide conditions today....