To some extend it does. At least for young-ish people. I don't remember exactly who did the reseach, but there was an oler psychological study on children who loved to draw. They painted every now and then and had fun doing it. Then their parents/supervisors started giving them rewards for painting, ranging from new paint to sweets to money. The painting frequency of these children increased and so did their art quality (of course, since they basically practised more), but their fun diminished and one day they would view painting as some "work" they did to earn a reward.
I think the bigger danger is the one that is ever present in gamification of anything - you end up incentivizing the markers of your goals and not the goals itself. It may sound weird to talk about the dangers of gamification in a game, but I think it is relevant when we are talking about creating game incentives in areas of the game which previously lay outside the mechanics of the game.
First, to background on this, some may not know that for a while there, there was this fad where you were supposed to incentivize work by making it more like play. Lots of people tried, but except in a very few areas, there hasn't been a lot of success. Mostly what they discovered was what was already known from rewards programs - people start gaming the system to maximize the collection of rewards for the minimum amount of work. The most spectacular failure recently was the Wells Fargo scandal where Wells Fargo had a rewards structure in place that incentivized the creation of new accounts, so in response their agents began to manufacture new accounts - entirely without the permission of the person to whom the money was owed.
Designers that give story or role-play rewards for objective story events (rather than subjective story-teller impressions) are hoping to incentivize story creation and role-play. But I strongly suspect that these rewards will mostly just incentivize the checking off of the markers of story, without necessarily encouraging building a strong and effective story structure. In other words, you will motivate players to pursue quantity over quality, and to treat role-playing like a sort of level treadmill to grind - "Let's see how efficiently and quickly I can check off the things on my list of things to do." The results are likely to meet the wording of the rules, but not the spirit of them, and the overall story is likely to resemble 'X-Files' or 'Lost' or 'The Force Awakens' or 'Twin Peaks' or 'Sherlock' or any number of similar post-modern plotted story lines where you make up the event or force your plot through some hoop and then struggle to try to explain it affords. The individual scenes might be effective depending on the skill of the players, but if you step back from them a little, they signify nothing and don't support a good story as a whole.
Now of course, you cannot mirror this study 1:1 on roleplay systems who reward good roleplay in a mechanical fashion. But such a system has to be designed carefully or you may have effects that you don't intend to. Let's say, some imaginary system gives your character one "quirk" and one "flaw". Roleplaying one of these two (or both) will give you some benefit. I guess that some players, especially those who hunt for benefits anyway will play exactly those two manners over and over again. They don't have to invest in playing another manner or think about their character's emotional, social etc. background in a given situation. Such a character can quickly descend into stereotype territory if you, as a GM, are not cautious enough.
Exactly. Actually, the real problem here is less likely to be a stereotype per se (since RPGs lend themselves to stereotypes) but the problem of trope decay in all of its various forms. The real motivation is out of story and mechanical. So the reason for the character to behave that way in character not only gets lost, but the mannerism becomes more and more clownish and loutish as the button is repeatedly pushed to get the reward. I believe the technical term is Flanderization. What had been simply a characters quirks become the whole character. Characters in a sitcom that are known as 'dumb' tend to get dumber and dumber the longer they are written, until they are not only a stereotype, they become a parody of their original character. You end up with flat unnuanced illogical often ridiculous characters that behave according to an increasingly rigid and extreme simple paradigm of what the character is known for. Their motivation decays over time until they are basically just parroting sayings and actions that they are known for.
A comparison might be the initial scene chewing by which we are introduced to Darth Vader and Kylo Ryn. Both villains are searching for something. We are initially introduced to Darth Vader holding aloft a prisoner and choking him to death. We are initially introduced to Kylo Ryn displaying similar villain badassery. But then Darth Vader says something that follows from his clear motivation, "Commander, tear this ship apart and find those plans. And bring me the passengers, I want them alive." But Kylo Ryn faced with the exact same problem says something that is not sensible from his in story motivation, and is sensible only if you view the purpose of the scene as 'display in story how evil you are'. It's as if the creator of the character had on his character sheet, "Gain 1 XP for mercilessly killing someone. Gain 1 XP for destroying an inanimate object." So instead of operating according to his stated character motivation - "I need to find this map", he starts killing the very people that could help him find the map, and destroying the very objects that could potentially be holding the map he's looking for. Hey, what about the map? This is frankly bad story telling, and while it makes for good scene chewing and I bet 90% of the audience just went with it, if you try to build a long story on a series of plot holes like that, you end up with something that if you step back from it is a mess.
Compare also with this awesome essay on the problems we see persistently in writing fiction these days:
http://www.tor.com/2017/01/18/sherlock-and-the-problem-with-plot-twists/