I would think forcing people to better appreciate how probabilities actually work is better than avoiding the topic entirely. Learning that 70% is not certain is important in lots of contexts. 30% probability events happen all the time…

Data illiteracy, not the existence of a forecast, is the core problem.