I've used both in the past. I also wonder if one is preferred over the other. I've usually done the later (without regex) due to simplicity and also that I do not agree with the Regex expressions that Checkpoint feels are best practice in sk165094.
The reason for that belief is because there is no termination of the expression. That is to say, if I'm only looking for example.com in the URL or SNI, then what is stop the regex from matching on www.example.com.bad.domain.com. The regex expression \.example.\com will match on that domain even though it could be a bad site.
I really think the regex should at least have a terminator on it, like \.example.\com(\/|$) which may work better on SNI hosts. That at least makes it match for what you want and either a backslash after it or nothing else. Even that is not perfect as you could still match on an HTTP URL, something like http://badwebsite.com/www.example.com/.
Again, one of the reasons I still favor the non-regex version of these filters.