This is the eternal dilemma of trying to determine whether a given email is likely valid. If you use a highly-specific regular expression, you get false negatives when a new top-level domain is introduced. If you use a more generic regular expression, you risk false positives. Your best option for this probably boils down to which of those risks you are more comfortable with.
If you look at the regular expression in the Community Question Library, you should see the top-level domains that it already supports:
If you want to keep this regular expression fairly specific, you could add your new top-level domains to this list.
But it sounds like you have a lot of domains in mind. Perhaps this warrants making the regular expression more forgiving. Near the end of the current regex's domains, you should see this:
That allows for any two-letter top-level domain to get through. That could be replaced with something like this to allow for any top-level domain that is two or more letters: