Yes. When you experience multiple unlikely failures the rule book goes out of the window. Difficult to plan for, if you need to go a long way down a chain of “what if?” to get there.
Bear in mind why we use approach minima, which is to give a very high probability, close to certainty, that you’re not going to hit anything at minimums given all of the worst-case inaccuracies in systems performance. If the alternative is crashing/ditching, then substantially reducing those minima or keeping going until you see something is a valid option, especially on a precision approach or RNAV/GPS based one. Yes, you wouldn’t want to do this as a normal procedure, as statistics will eventually catch up with you, but as a one-off in extremis, it’s probably not as risk-bearing as you might envisage.