To what extent? Would such data meet the expectations of peer-reviewed journals? What is the difference of influence between digital and non-digital communication?
The answer is yes in my mind to the extent that rigurosity in recording and handling of these data is observed. I do not know how journals would react to studies based on non-traditional field work. However, Hispanic linguistics has only to lose if for instance objected to research questions being responded on the basis of 10,20, 50 or 100 hours of recording of chilean live television. Same for peruvian, andalucian, or spanglish TV or radio broadcasting.
I see no reason to object data from these sources. These are means through which language users express themselves spontaneously and authentically. I am all for using digital communications as data sources.