He's NOT the messiah!
Sorry, couldn't resist. However, his message is not a universal truth. We still need to throw stuff away, even without GDPR. Many instruments generate such a massive load of data you cannot possibly store all of it, even with cheap storage. I work both on data from astronomical instruments and nuclear accelerators, and the data rates of the new generation of instruments, like SKA (but even for existing LOFAR) means we must reduce the data at the source, and throw most of the raw data away. This is on the one hand anathema to scientists, but it has to be done. Most collisions in accelerators yield nothing of interest, and can be discarded. It is just the rare events that need to be stored. So he is right that we want more data, but we don't want to store indiscriminately, or we will drown in a see of uninteresting stuff.
He is of course right that we need more processing grunt (and storage), but that is pointing out the bleeding obvious