Reading is a Developer's Superpower
People who are into mathy-sciencey fields are generally not known for being good readers. There's more written information on the internet today than there ever has been, and software developers are great at skimming through lots of information and finding exactly what they need to solve the specific problem they're facing at that point in time. This skill is certainly valuable; being able to pick the offending code from a stack trace is an essential skill for efficient, independent problem solvers. In my own personal growth though, I've noticed that this skill is at odds with my general reading comprehension ability. Not to say that both can't exist together, but at times I feel like a child learning to play golf and baseball at the same time: the swings use roughly similar muscles, but they are far from complementary skills. I would like to make the case for more software developers to read long texts, and read a lot. Learning from written text is a valuable skill that unfortunately seems to be losing mindshare among our profession.
Shortcuts abound
I started my career in software development following the traditional path: I went to college and got a degree in computer science. You would think that such an endeavor would require a lot of learning from things like textbooks, academic papers, and so on, but unfortunately you would be wrong - not that you can't learn from those things in university (and some universities may be better at pushing those things than others), but there are usually easier ways to learn exactly as much as you need to complete an assignment and move on to something else. Very few programming courses taken in school are complicated enough that you can't find a video online walking you through the concept. I used these a lot, and in the worst way possible. Sitting through an entire 20 minute video was too much of a commitment, so I'd scrub through until I heard a few words that sounded loosely like what I needed, listen to about three minutes, and then close the tab and move on. If I needed more information, I would most likely end up sourcing from an entirely different video. I didn't really "learn" much; by the end of many of my courses, I had a lot of disconnected ideas about topics such as networking, algorithms, security, and so on in my back pocket. These were far from worthless! As I've grown, I've been able to fill in a lot of the gaps between these ideas. But it took a long time for many of them to become connected enough to be considered "workable knowledge".
"Islands of Knowledge"
One problem with the "Islands of Knowledge" is that they usually lack any sort of conceptual integrity since they come from many different sources. Clips of YouTube videos, comments on a Reddit thread, and the tl;dr at the bottom of a blog post may all be "technically correct", but the context built up by each respective author gets lost when information is poached as needed instead of taken as a whole. You may find the answer, but that's it. At best, you lose a bit of context and history that isn't immediately relevant. At worst, you missed enough nuance that your understanding is actually wrong. Reading the entire body of a work gives us a window into the mind of the author and lets us better judge the end result.
As much as we probably hate to admit it, there aren't a lot of "facts" when it comes to writing software. For larger problems, things need to be broken down and reasoned through to reach a subjective "best" solution. Trying to stitch together a bunch of "Islands" only gives the illusion of working through that process. For example, reading the summary of "How we scaled to 10 billion requests per second with Kafka" may give you some talking points and a basic overview of important terms, but you didn't learn things like how the author faced the problem, the questions they asked along the way, the things that did or didn't work and why they did or didn't work, and so on. Synthesizing the sparknotes of that article with one written by someone working at a different company on a service designed for only 100 requests a second will probably not yield very useful results; there will definitely be some overlap in the "Scaling with Kafka" Venn Diagram between the two (the terms they use may match for example), but also a lot of blank space. When it comes time to discuss this with someone, you'll have to connect these islands together yourself on the fly, but since you didn't get as much context as the authors provided, you have no way of knowing how close together the conclusions of those two pieces really are. You would certainly still have to make some jumps in reasoning if you read more than just summaries and snippets, but the distance you'll have to cover will be much greater if you didn't. And greater leaps with greater uncertainty are rarely a recipe for success.
Reading Metaskills
Reading complete pieces gives you a more complete picture of the information presented than simply reading the summary or skimming it, but there are other skills that get trained too. I've found that stretching myself to read longer and more complex documents helps me keep longer and more complex ideas in my working memory while in discussions. Reading and comprehending lets you practice slowly so that when it's time for the discussion, the knowledge has already been worked and compressed to the point that it's no longer difficult to recall or reason about on the fly. Your mileage may vary but I've got to imagine that this skill of "working and compression" is one that's improved by intentional repetition. Another oft-cited benefit is increasing your vocabulary. Having a larger vocabulary allows you to use your words much more precisely. I've found this to be immensely useful when explaining concepts to people, or writing documentation, or sharing opinions on topics I've read about, and so on.
Wrap-Up
We've been conditioned to search for information in ways that are unfortunately detrimental to our ability to read and comprehend longer forms of writing. We learn, but in a disjointed, inorganic fashion. This cycle is self-reinforcing, and while we obviously will not lose our ability to read more comprehensively, we've grown averse to it. But there's good news: it's never too late to start linking your "Islands of Knowledge" together. Next time you have to gather some knowledge about a subject, start small - pick one blog, article, or paper that you would have poached information from before and read it all the way through, even if it feels like a waste of time. You don't have to take notes or write your own summary; just read it to from start to finish. You will likely learn something you didn't set out to learn, bringing a few disparate Islands just a bit closer together in the process. If you keep at it, I can assure you this small investment will pay dividends in short order.