EDIT: Or, to be fair to the authors, other factors besides relative reading ability, as well as as accounting for the other 20% of the gender gap in math-intensive fields in general, also explain the gender distribution between particular math-intensive fields.
I do agree that the aforementioned fields are not necessarily math-intensive — except investment professional in certain cases.
An accountant or tax adviser needs reasonable applied math skills, but isn't likely to need to know the difference between a cardinality and a cartesian product.
Likewise, most working programmers don't need to know the difference in their day-to-day work - although they might need to in order to pass a CS degree as a gateway to getting hired.
The classes are at least initially perceived as easy for most people, but while some find them fun other find it horribly boring. So, my guess it’s less about aptitude than liking the kind of work involved.