This happens all the time - they are fish, and they need to eat to stay alive, so it isn't too surprising that they forget the sting of the hook after a few tens of minutes. I have numerous (probably dozens) of cases where I've caught the same fish multiple times over the course of days or months or even years, and in some cases, the fish was caught by a fellow PAFFer. So for all the times that I've been able to ID a fish again, there are probably many more where I didn't, either because I just missed the fact that they were the same fish, or a PAFFer didn't post the fish. Not downplaying the OP accomplishment - I think it is neat to recognize a repeat catch and is a testament to the benefit of C&R, and also some evidence that photographing a fish doesn't kill it ;-)
It's much easier to correlate the same fish if there is something unique; for instance, tigers are pretty easy to compare, because they are relatively scarce (i.e. small sample size). Other variables are if a fish is a home body and you happen to catch it at the same spot a couple of times. If it is memorable the first time, and you catch a similar fish there a second time, then I'm more likely to compare the fish and find they were a match. Or if there is a unique spot pattern, or marking, or wound, those tend to register. The nondescript "normal" trout are the ones that I'm likely to not register as being caught more than once. Or, if enough time goes by, a two-year old fish that survives to be a four-year old fish might be different enough in body shape to not register. Also, the probability of a fish surviving for four years is low, so I first have to catch it as a two-year old, then it has to survive another two years, then I have to catch it as a four-year old, and I have to make the correlation between the two fish, which all sums up to being a low-percentage game.. I also think this is less likely to happen when the fish density in a stream is high - you just have a much higher sample size to catch from (e.g. Spring Creek). When you are fishing a small stream, and you have a sample size of say two fish in a pool (one being a pool boss), you have a much higher probability of catching the pool boss the first time around, which will leave an impression, and when you fish the pool again, you have an equal chance of catching that fish, and when you do, you'll remember it.