If your argument is that there is a blurry line between recreation and work when you love your job, then sure. Maybe the average person will believe that in some abstract intellectual sort of way. But I don't think you'll convince any mathematicians that their entire job is recreation. By analogy, I don't think you'd convince any world-class musicians that their entire job is recreation when they're doing music drills and rehearsing for ten hours a day.
puzzles = the most fun part of math
nor
puzzles = the best, most significant part of math.
is always true. Sometimes, sure but the entire math terrain is complex and shouldn't be reduced to pure puzzles, pure progressions of abstraction or any other simplistic view.
It is a recreation, of a kind, that shouldn't be taken for granted even after doing rote derivations for ten hours a day. If you enjoy it that ten hours is rarely enough.
That's how I interpreted the OP.
I become interested in mathematics at a young age. While puzzles and such have sort-of interested me, the truly amazing part of math has always been it's ability to open up more abstract ways of thinking - new and exciting ways to unify ideas. (I read Gardner as a kid and got rather little of him. His puzzles go so quickly from simple games to really hard math that it's easy to get lost. I honestly would see him as a terrible introduction to math for kids since a kid reading him will hit quickly "why can't I solve this even slightly" moments, which isn't a way to encourage someone).
A lot of puzzles to my mind seem to have "trick" solutions. They're nice for increasing one's ability to manipulate symbols but they aren't really introducing new worlds into one's thinking.
I enjoy abstract algebra, functional analysis and differential geometry more than number theory, which is the branch of math closest to ordinary puzzle solving. On the other hand, category theory seems more like accounting to me but clearly category theory has yielded results that have advanced many parts of math and computer science.
I can certainly acknowledge that others might have different tastes and the development of math has certainly has come through different people developing their particular interests rather than there being a simple "fun"/"not-fun" division.