That's why I consider veganism an experimental sort of diet.
I mention veganism because vegetarianism is even LESS problematic; a lot of the stuff you get from meat but not plants, you can still get trivially from eggs and dairy and such.
I'm aware we don't know absolutely everything about nutrition, and that's why I still consider veganism to be a somewhat experimental thing. However, we do know quite a bit about it, especially compared to what we used to, and as far as I can tell and have been able to find out, it can work and can be healthy.
The kind of uncertainty you're bringing up is one reason I probably would never become a vegan myself, but that's part of the reason why I respect them in the first place, as long as they understand that risk, as much as it may exist.
In fact, I think that's one reason why it's good to have vegetarians/vegans around. There's essentially no better way to study whether or not it is healthy, and it's getting popular enough now where, 20 years from now, we're likely to have enough case data to draw stronger conclusions, if not absolute ones.
Really, you can say "do you want to mess with something you don't understand?" but that you can apply to any seriously dietary change, and you can only push the goalposts back so far before saying "yeah, go ahead and try it". We know enough about nutrition now to know that it isn't exactly going to kill you, provided you know what you're doing, will probably be reasonably healthy, and like I said, enough people are doing it that by the time they're old, we'll likely have our answers.
After all, when it comes to complicated stuff like this, at some point you have to bite the bullet and experiment in order to get the data you need.