Of course, I am not suggesting anything. The springs are strong but naturally nowhere near that strong. On the other hand, the adapter protrusion under the spring might of course cause that (and now that I took a second look, I can see the end of the spring "printed" in the middle of the damaged area).
I don't know how much a cell could expand without breaking. Probably not much. But that kind of damage requires quite a lot of brute force, and since you didn't apply it, I didn't have any other explanation how this could happen. I know I used rather much force when I tried my light with that Redilast 2900 cell and managed to screw the tailcap to the end. Even then it didn't do anything to the negative end of the battery. Neither to the light - it seems to be well built. Some of the longest protected cells were too long (at least when the light was first released), but most work fine. And even that Redilast cell worked in some other people lights. Factory tolerances, I guess (either the light, or the battery). Maybe you got a light that is out of those tolerances but managed to pass the QC. I believe Xyber will sort it out.
And yes, I wouldn't use that battery any more, or store without a protective bag in my house, either. Some quick tests like the remaining charge/voltage might be interesting though.