Maybe there's a transitive relationship between the environment before and during transcription and the resulting RNA that in turn can affect its environment. It would be difficult to identify in vivo because the environment is already synchronized with the changes that may have been introduced from the RNA, but in vitro maybe the effects of the altered RNA would become apparent in an unmodified environment. Vice versa, if you could measure enough samples of RNA from the same organism and source DNA I wonder if there'd be a quantifiable difference when the environment was subject to the same sort of changes that occur when memories are formed and an unmodified environment.

Really cool research!