As Tucumseh said, quite a lot can be done with modelling but if it had been me making the decision, I would have insisted on LOTS of environmental chamber testing, including (if access could be gained) use of a Climactic Wind Tunnel. Beyond that, the process should have been incremental testing and "learning through experience" on training, test and operational flying, seeking to build a body of evidence over a period of time.
The problem with the "all weather" capability as marketed at the outset of the WK programme is that it wasn't something the manufacturer could actually deliver, and so they ended up trying to prove a capability that was in all likelihood going to result in the loss of the aircraft. The pressure to deliver may have been contractual i.e. the MOD insisting on it being tested to their satisfaction. It's bad enough being iced in a manned light aircraft, let alone a UAV that has positive 'pilot' control, and of course, WK is "point and click" so the operators could not intervene i.e. try to find a break in the clouds and get the hell out. Remember, the programme had already suffered an airframe loss due to water ingress / blockage in the pitot tube.
To anyone with a decent amount of UAV experience, putting the aircraft into these conditions was a pretty foolish thing to do and it suggests that those making the decision to do so either did not appreciate the risk, or did so but pressed on regardless.