This may be a somewhat philosophical question, but here goes.
Wikipedia claims that several nuclides (e.g. hydrogen-5) have half-lives shorter than $10^{-22}$ seconds. This is on the same order of magnitude as the amount of time that it takes light to cross a nuclear diameter - i.e., the shortest possible time that (it seems to me that) it even makes any sense to talk about the logical possibility of a bound state.
Given this, is it really meaningful to call something like hydrogen-5 an "atomic nucleus", as opposed to just "one proton and four neutrons that happen to momentarily pass very close by each other, interact via the strong interaction, and have the proton and two of the neutrons fuse into a tritium (hydrogen-3) nucleus while the other two neutrons go on their merry way"?
Clearly once we get to nuclides with half-lives in the seconds and minutes, it becomes useful to think of them as metastable bound states (and indeed, for the purpose of applications like atomic or molecular physics or chemistry, to abstract them out to new "primitive particles"). But is there any kind of sharp (or even semi-sharp) notion of a "nucleus" for the incredibly unstable nucleides - e.g. any specific phenomenological signatures that differs from just colliding a bunch of protons and neutrons together? Or does a concept like "a hydrogen-5 nucleus" really just summarize "the complicated scattering process that ensues when we collide together one proton and four neutrons"?
