Mobile phone screens are good at accomplishing discrete activities with defined workflows. Indeed, they're much better than a desktop computer for things like:
Paying for gas.
Depositing a check.
Scanning groceries at the market.
Checking in to a flight.
Buying something from an e-commerce shop.
They're also amazingly good at a certain kind of reading—the lean-back kind, where you're meant to consume content in a linear fashion. Indeed, the (relatively) small screen distills reading to its purest form—one where you can't even see much more than the paragraph you're currently reading.
Phones truly excel at providing content for what Ben Thompson calls "the available time around intent," those moments when you're waiting in line or riding the train or any of the million other situations in which we have a few minutes to kill while waiting for something.
The time around intent is perfect for the feed. Content streams by and we dip in and out as something catches our eye. It's all short enough to be consumed before the next register opens up.
There is value in the feed. It's an important part of a content everywhere approach to transmedia storytelling. Sometimes the tweet version of a research report is all someone really needs.
But leaning back in the time around intent is not how research is done. Research is a sprawling mess. It needs more real estate than a phone provides.
To put the point into more modern vocabulary: where the modern web focuses on the best way to display a single text, these earlier systems focused on providing a dashboard that highlights the relationship between a given text and the other elements in the collection.