Not sure if that's the falsifiability you meant, but a scientific theory _must_ be falsifiable by definition. As in -- there must be a way to setup an experiment with a possible outcome that would prove this theory to be wrong. It is impossible to disprove a theory that "invisible pink unicorns exist" -- because for every suggested experiment you can just say "well, they are invisible, so we can't see them". Of course falsifiability is not the worst problem with this particular theory (how can something be pink and invisible bothers me more for example), but it is a necessary condition any scientific theory must meet.
I brought up falsifiability casually because it is an important corner stone for my understanding of the scientific methodology and keeps scientists from proclaiming dogmas (or at least it should). It is also an acknowledgement that our current theories mostly reflect our current understanding and may be prone to change in the feature.
