I'm not quite sure what he Gelernter is getting at. It seems that he thinks the temporal character of information is the most crital one, and thus we can tie all information together with this unifying aspect.
I don't believe that is true.
IMHO much more likely, the web will fragment and map onto objects and spaces in a manner that best reflects the utility of the interaction. Time will be just one aspect in a number of them that will enable relavent information to be drawn up.
The most timely information has the least context.
Herp. Derp. So here's the thing: time is a dimension. Just because you are not traveling through it does not mean you are ignoring it. To the contrary, by looking for "time streams" you are forcing yourself to rigidly march through information - rather than seeing things as part of a whole, you are seeing things as a linear progression. The timeliness of information is absolutely important, but so is the perspective within that time. His argument is basically the same as saying "you have Twitter, you don't need CNN" or "you have spotify, you don't need iTunes." Perhaps this reflects my own prejudices (I don't follow Twitter, I don't listen to Spotify, and I go to places, not times) but I choose to regard every dimension of the information available to me. Why listen to the radio when I can find whatever song I want? In justification, the dude cites infinite scrolling - which people always use instead of search, right? I dunno. It takes some stones to suggest you "predicted" the "world wide web" in 1991, a full ten years after Neuromancer was on the NYT bestsellers' list.
Time is spatial. Perspective is implicit of a point of view. A point of view implicit of an observer. An observer is implicit of a field of vision. A field of vision is an affordance of visible objects. So a perspective view is an aggregation of objects near and far at a given point of time. Actually WWW was predicted in the middle ages.