Therefore his predictions are not that much original research/predictions, but rather a good compilation and clever choice of ideas which had appeared earlier in various sci papers. I suspect his quite striking description of "Kindle reader" from 60s was based on some previously read paper/article. Still, it's a very very good prediction as for 60's.
To add to his "predictions": Golem XIV is a philosophical desciption of technological singularity (from 80's), and his Summa Technologiae (from 60's) tackles a ton of then veeery novel stuff, like Machine Learning (a precise description even for today's standards), SETI research, VR, AI, and all of that in a "scientific" way, i.e. concreete and directly based on then published papers, and not simply made as an effect of extrapolation of then current trends (sort-a what Asimov liked doing)