Actually playing fast and loose with statistics has its shortcomings. I dont think that a single sports statmeister can defend the significance of his stats with statistical rigor. What distribution do the population statistics follow? Never heard an answer by any "sports statistician" to that one.
Calling a game "insane" does not mean it can be removed from the sample population, especially one with a small number of games(samples) overall. Secondly, last years data came against a larger number of opposing defenses, and is therefore even more reliable in a general sense than the 3X more samples would indicate. Third, any pro scout would tell you that FG% depends on the defense and the type of offense that is used to break down that defense. Are these last second, heavily defended shots? Since the Kings offense of last year is more similar to the Suns offense than the other teams that House played for, it is probably a substantially better indicator of House's expected performance in the suns offense. Fourth, if a player is developing(a sliding mean), the more recent performances(not the first two years of his career) are probably more indicative of where that player is now. If he is learning a new offensive system or playing with new teammates, it make take some time for the adjustments.
Because of these concerns, we are all playing fast and loose with sports statistics. Its just that failure to consider these influencing factors makes the stats faster and looser. In actuality, the performance of athletes should be broken into a number of smaller populations, considering the influences that change the population statistics. Averaging across all populations is especially ineffective if a player has played in a number of different teams/systems, like House. So yes, there is a difference between the 27 shots this year and the 97 last. I would even say that with Barbosa or Nash in the game, House would get more open shots since Barbosa's (or Nash's) presence will cause Houses' defender to stay closer to the lane.
If one says I don't like the ability of House to take a shot while defended, I tend to agree, he's obviously a better catch and shoot guy. But "House is not a good shooter" is a trivialized evaluation.