Your question seems like it could be interpreted in a few ways, but the law of large numbers is going to have a hard time identifying someone who is good enough at gambling to make a living (as opposed to someone that turns out to simply be lucky, which the law of large numbers suggests is likely to occur more often than never).
Additionally, the amount of winnings depends on the size of the stake, and so a bettor would need to already have a decent amount of money to be able to win enough to live on, and so a strictly professional sports gambler might not be feasible (in that another source of income or independent wealth is probably necessary to win enough that one could live on the winnings alone).
I'm not aware of any empirical cases, but that's not surprising as the majority of sports betting in the U.S. is likely to have been illegal until recently, and a consistent winner may not want to draw attention to themselves either way.
Interpretation 1: Is there or has there been anyone whose net return from sports betting has been positive?
Almost certainly.
This is "law of large numbers" territory. With the number of games played and the variety of types of bets that can be placed (winner, point spread, tournament brackets, fantasy leagues, etc.) there have been an enormous number of wagers placed among a huge number of people. For this interpretation to be correct, only one of those people needs to win more than they lost.
The time horizon being discussed is very important. Someone who wins consistently for 20 years may have "made a living" off of it, but a string of losses thereafter could still easily drive their long-term returns into negative numbers. Someone could, for example, just be lucky enough on which bets they placed and in what amounts that they happened to win money.
Interpretation 2: Is there or has there been someone who was so skilled at predicting outcomes in sports that they actually have a positive expected value from any given dollar wagered?
Soft maybe. This is hard to evaluate because the "skill" of the choices such a bettor would make is essentially impossible to differentiate from the lucky gambler above.
An important difference between sports betting and lottery tickets or casino betting is that sports bets are not based on well-defined rules (like 5 numbers being independently drawn). Instead, bookmakers set odds, and they do so not based on what they think the outcome will actually be but rather in order to attract bets in the correct proportions that the pool will be able to pay out to the winners, plus a margin for the bookmaker to keep.
In this case it's less about predicting outcomes successfully than it is doing so with uncommon information, such that the odds allow for winnings based on that information that are enough to cover living expenses while leaving enough cash to stake in the next bet. Whether or not the odds allow for that information to be profitable is a factor totally outside of the bettor's control.
Interpretation 3: Is there or has there been someone who is sufficiently skilled in the institution of sports betting that they can have a positive expected return on any given dollar wagered?
Perhaps, but the time horizon is again very important here.
A gambler who can manage a true expected positive return is far more likely to exploit arbitrage opportunities across multiple bookmakers (offering odds for different outcomes that can "balance out"), and so hedge their bets effectively, instead of to be great at predicting outcomes of sporting events. This is obviously heavily dependent on decent math skills and the environment in which the betting takes place as opposed to being good at betting on sports, specifically. I would not expect such opportunities to exist steadily over any given stretch of time.
Nonetheless, if someone only bets in these situations and places large enough bets to live off of the net winnings, then the net expected value of the actual bets could easily be positive.
Interpretation 4: Is there or has there been anyone capable of making more informed bets than others such that they can derive enough winnings to live on?
I suspect that this is the closest to your intended meaning, and the answer to this is moderately likely to be yes, provided that we are not limited to publicly available information. Knowledge of things like rigged games obviously has very high value for a bettor (though in that case they're not exactly betting...).
But there are other bits of valuable information an insider could gain that might give an edge over the average bettor, such as knowledge of a putative injury to a player or an upcoming suspension, which might allow for favorable bets to be placed before that information becomes public (if it ever does) and is then incorporated into other bettor's decisions and the odds offered by bookmakers. It's not crazy to imagine that a well-placed or well-connected person might be able to gain enough information to generally win bets.
It's probably also possible, if not necessarily feasible, to seed the environment with information (true or not) in order to manipulate odds offered in the bettor's favor.
Final Note
I would think that the people that understand sports betting markets well enough to have an expected positive return on "honest" bets are very, very likely to be bookmakers instead of bettors. Bettors have to take odds as they find them, and so (aside from inter-bookmaker arbitrage) can't necessarily leverage information into sufficient winnings even if their knowledge is guaranteed.
