My reason for thinking the flashing-words-in-a-window thing will increase reading speed is that there's already software that does it, which has been confirmed to work. Most people don't use them more than briefly, though, because of the inconvenience.
Software that does it (google-friendly phrase: "rapid serial visual presentation") has been available on all the major platforms (Windows, Mac, Linux) for at least 15 years. I know from personal experience that on Windows and Linux, the software worked reliably and without any annoying glitches or performance problems. Yet hardly anyone uses it more than briefly. What exactly is this confirmation that you speak of that would move the expected value back to anywhere near where it was before taking into account that hardly anyone in the last 15 years has used it more than briefly?
"Hardly anyone": am I wrong about that? Can you name two people who do not have a large financial or academic-prestige stake in it who have ever used it more than briefly?
It is not enough to show that it has benefits. You also have to show that it does not have major drawbacks that cancel out the benefits. I have a specific drawback in mind that with probability .88 cancels out any benefits for the vast majority of potential users. It is not ignorance on the part of computer users or inconvenience. My experience with using it on Windows 95 and Linux was that although it would be more convenient if softwares like browsers and operating systems were better integrated with it, the inconvenience of using it is not particularly high especially after one takes into account that the user can reserve its use for longish passages while handling short passages the ordinary way. (Do not recall the name of the Windows RSVP software I tried. The Linux software was call sview.)
I have a specific drawback in mind that with probability .88 cancels out any benefits for the vast majority of potential users.
What drawback did you have in mind? I'm confused why you would allude to this without specifying what it was, and curious as to whether it's one I've already thought of and tried to mitigate, one that I've thought of but can't mitigate, or one that I haven't thought of.
Recently, I've collapsed (ascended?) down/up a meta-learning death spiral -- doing a lot less of reading actual informative content, than figuring out how to manage and acquire such content (as well as completely ignoring the antidote). In other words, I've been taking notes on taking notes. And now, I'm looking for your notes on notes for notes.
What kind of scientific knowledge, techniques, and resources do we have right now in the way of information management? How would one efficiently extract useful information possible out of a single pass of the source? The second pass?
The answers may depend on the media, and the media might not be readily apparent. Example: Edward Boyden, Assistant Professor at the MIT Media Lab, recommends recording in a notebook every conversation you ever have with other people. And how do you prepare yourself for the serendipity of a walk downtown? I know I'm more likely to regret not having a notebook on hand than spending the time to bring one along.
I'll conglomerate what I remember seeing on the N-Back Mailing List and in general: I sincerely apologize for my lack of citation.
Notes
(I've also heard a handful of individuals claim that SRS has helped them "internalize" certain behaviors, or maybe patterns of thought, like Non-Violent Comunication or Bayes Theorem... any takers on this?)