"Probabilistic fallacies" aren't "formal fallacies" because formal fallacies have a very narrow definition: they are fallacies of formal logic -- be it FOL, SOL, infinitary logics, modal logics, etc. -- e.g. illegal "moves" that break particular rule-sets.
Probabilistic fallacies deal with incorrect conclusions drawn from looking at data. Or, in other cases, misapplications (or misunderstandings) of mathematical laws. These kinds of mistakes have nothing to do with "formal logic."
Sure, but if we're going to have a list of fallacies to help people identify fallacious thinking (and use it as a blunt weapon in internet slagging matches) then we might as well include probabilistic fallacies in that list. They are basic mistakes in reasoning just as formal logic fallacies are. They can always be listed under their own category.
Anyway the calculus of probabilities is a logic. It's only for historical reasons that this is not more widely recognised (and its unfortunate association with statistics).
Edit: the website itself doesn't say anything about "formal fallacies", just "logical fallacies". Which seems to be used in an loose manner that should admit probabilistic fallacies in the set of "logic fallacies".
One thing the site could do is extend the "periodic table" metaphor to at least "color" the differences between the types of fallacies involved and give slightly different styling to formal debate fallacies versus probability fallacies.
Probabilistic fallacies deal with incorrect conclusions drawn from looking at data. Or, in other cases, misapplications (or misunderstandings) of mathematical laws. These kinds of mistakes have nothing to do with "formal logic."