I think science has shown with some goldfish that once they're moved from a small environment, once placed in a larger one they grow at a rapid rate, but I couldn't cite where exactly that is from.
I've heard this too, including being told about while at university by people researching fish farming. Fish grow continuously, although faster when young. They are growing in size until the day they die. This is obviously different to birds and mammals, which tend to get to a certain size at sexual maturity, then stop growing. Reptiles, on the other hand, grow continuously as well. Anyway, this means that if a fish is transferred to an environment where it can only grow slowly (e.g., little food) to one where it can grow rapidly (e.g., more food) it will "seem" to start growing again because its growth rate will increase dramatically.
With cyprinids -- like goldfish -- and some salmonids -- like trout -- there is a feedback system between individuals. I'm not sure if it has evolved for a good reason, or is merely an accidental thing that ends up being significant in fish farms and aquaria. Anyway, the biggest fish produces more chemicals (called metabolites) than the smaller fish, and these chemicals have more effect on the smaller fishes than the bigger ones. They slow down growth. So the bigger fish keeps getting bigger, but the smaller fish grow much more slowly. For fish farmers, the problem is ending up with a pond of fishes of which one specimen is big and all the others are small. What farmers want is lots of similar sized fishes they can sell easily. Tilapia don't do this, but trout and carp do (if I recall correctly).
The "just so" story is that this mechanism allows the dominant fish to stay the biggest, which is a good thing for the dominant fish. In the wild of course the sheer volumes of water dilute these metabolites so they don't usually have a very great effect, so a school of trout or carp can all reach a good size, which is important for their social behaviour and safety-in-numbers (a school of fish consisting of one big fish and a hundred tiny fish would be pointless!). But there might be just enough effect that it has a subtle, but significant, impact over time, giving the slightly bigger dominant fish access to more food, better mates, better hiding places, etc. In a pond (or aquarium) the dilution factor is so small that the metabolites build up and slow down growth much more significantly.
A single goldfish (say) kept in a very small bowl or tank, and given few water changes, absorbs its own metabolites, and these have the effect of slowing down its growth rate, as if it was sharing a pond with another big goldfish. Hence goldfish seeming to "grow" to the size of the tank they're kept in, especially if they are kept really badly.
As for puffers, who could say? No one really knows enough about these fish, and I don't think much research will have been done one them with regards to stunting, and larger environments.
Agreed, I'm not aware of any such work. My gut feeling is that they don't stunt in the same way as salmonids or cyprinids because they aren't schooling fish and they are very distantly related to the fishes for which the effect has been described. On the other hand, I have no doubt that puffers kept in too-small tanks get unhealthy, and under such conditions their growth rate may be compromised and certainly their sensitivity to disease will be greater.
Cheers, Neale