At this writing, there is a 121 rating point difference between the mean Bullet and mean Blitz ratings, to the favor of Bullet - not entirely insignificant, even if less than 1 standard deviation.
Also, close examination of the right-hand tail of the Bullet versus Blitz ratings distribution curves also reveals the bias in favor of Bullet (e.g., the % of players between 1950 and 2050 in Bullet is 0.72%; whilst in Blitz, it is 0.40%).
Of course, none of this means that ratings are necessarily inflated. They may simply reflect the differing player populations (e.g., one does anecdotally tend to see higher rated players preferring Bullet).
So, the answer is ... maybe there's a problem, maybe not. 

I know know bullet chess isn't reall chess but I'm talking about ratings.
Ok, so everybody seems to whine about the system being inflated but I looked at the stats and it seems they're just like the other ones, they peak at around 1200 and fall on both sides. What's the problem?