No, I'm not advocating not being prepared. What I'm advocating is being prepared for likely events - as an example, in one organization I was in, there are those who advocated practicing pedal failures in the hover where the pedal would go to maximum pitch (left). There was no hydraulic system in the pedals, and no possible way for the pedals to go to maximum left pedal pitch in any of the failure modes. The problem was that by introducing this false failure into training, it made people miss the point that the problem failure in the hover was loss of tail rotor drive - people were going to spend time trying to figure out if they had a stuck maximum left pedal or a loss of tail rotor drive (am I rotating left or right?) when it was pretty clear that if you started rotating rapidly in the hover, you needed to get rid of the power quickly.
The other thing I'm advocating is that if a stuck collective really does happen, then you ought to be reporting it. If you have something wierd like this happen, somebody in the aircraft manufacturer and someone in the certification authority should be told. The Service Difficulty Reporting system is set up for exactly that.
The helicopters are designed and tested so that these things should not be able to occur - if they do and enough of them do, then something will be done about it.
As an example, I've just seen an accident report on a helicopter that has a collective hold down system that (naturally) got engaged at the wrong time. I'm not sure how this model of helicopter still had this feature, as it was mandated to be removed in several countries as it had been implicated in several accidents / incidents.
And there has to be a limit to what failures we teach people about. "let's look at the failed pitch link failure..."
And just to show the other side, at least one country has mandated no intentional hydraulics off flight in one model of helicopter because of the problems of controlling the machine - this is perhaps unwise, because how do you train people to handle that emergency (which, in the model of machine concerned, happened with a significant frequency)?