My 2 cents:
Since science is almost exclusively (there are some exceptions) linked to population, and science is the end-all/be-all "power source" for your civilization regardless of your motives (the biggest flaw of this game), this wonder has one of the most prominent effects on the game-play experience, even more so as it comes so early in a game of exponential growth. Other events/tactics/choices that have similar impacts are:
-getting 3+ free techs from huts.
-selecting Poland, Babylon, or Maya
-selecting Isabella and being first to find GBRX2
-getting Great Library, slinging Philo and getting NC before turn 55.
-reducing your game difficulty by a level or 2
Of those, the 1st is game-breaking luck (why I play w/o huts), the 2nd and 3rd are unbalanced game elements. Only the 4th is a strategic choice, although it's usually not an option on any level higher than Prince (well, about 50/50 shot on King, but I digress.) The difference between ToA and GL-slinging Philo is that the former is almost always an option up to and including Emperor level and about 70% likely on immortal. The optimistic view of this is that it's a great strategic choice that's very likely to pay off; the pessimistic view is that it's an unbalanced game element, much like worker-stealing and should be avoided.
Also, my experience at Deity level (admittedly limited; only 4 attempts but all achieving a victory condition) differs from that of a previous post regarding this wonder. I agree that the AI seems to overlook the importance of this wonder at all levels up to and including immortal, but not at deity. I've found that the AI seem to each want to achieve at least one wonder, with one or two civs hording as many as they can. Of my 4 deity games, this wonder went on turn 26, turn 29, turn 34, and turn 39 - all standard game speed. Furthermore, if all your efforts go to achieving this wonder (things like researching archery before pottery, which at deity means you're probably skipping religion entirely) and everything goes great with game settings (like having at least 3 tiles that are F+P=3 accessible and getting an early worker steal) you're going to finish the wonder between turn 33 and turn 37. So judging from my previous experience (which, again, is limited, and keep in mind the universal rule of advertising - "results may vary"), there's about a 25% chance of success IF all the variables work to your favor.