Heh, it's not that much of hard work as I happen to have many components/tools/capabilities already from years before. What was really tricky at times is that LaTeX isn't exactly built for what I'm trying to do – it allows authors to get nice typesetting under the assumption that the author might adjust the text to make it fit, but for automatic processing of a fixed source, one can run into problems for which there's no package or good solution available. There are other processing engines (PDF generator backends) of the kind used by big publishers or for reporting, and I have used some of them some time ago, but not that much, therefore why not try to get away with LaTeX another time, for now?
Obviously the URLs in the side margin are the most pressing open issue right now, because of them overflowing. For some strange reason, a minus is not considered a good place for a line break (maybe it's a special rule for URLs), but even with that, some might be just too long with consecutive characters alone. Up to now, we have collected three possible solutions for this: (1) manual hinting (hard breaks, but not nice in terms of automation because they have to be placed/introduced manually), or (2) like with the Peeragogy layout, have them as end-notes (for more space) via an LaTeX index, and there's also some solution in there to wrap/line-break them (but I think it doesn't work all the time, but has much better chances because of the larger space), or (3) another great solution I've figured out for my own projects but don't want to release it under CC0. There might be a (4) from the LaTeX package side I have to look into.
I'm very suspicious of link shortening, as it gives some other external entity total control over where the links redirect to (plus tracking plus what happens if the service disappears, but you can't update the shortened links that are printed on paper in a book), and in terms of domain identity/authority or URLs as ReST IDs, there would be no point in using URLs in the first place, as one could do a lookup from short (local and/or arbitrary?) identifiers to the long URL under the hood or even on paper with an index or something (like, along the lines of there's never a good reason for three levels of indirection, conceptually). So sure, shorteners might solve the immediate problem of long URLs conflicting with typesetting layout (by the way, print/paper medium is really not to be conflated with or great at immitating electronic hypertext media, and even if text is written for a rather primitive system like the Web, it can hardly work via a simple 1:1 translation, which is why print/paper too would need some innovation to improve it), but at the cost of rendering it's technical/navigational function more fragile.
In general, the Wraps tend to contain more URLs than the Handbook for the reason of pointing to many things and resources, and by being relatively short, they likely don't come with a lot of footnotes, so the Tufte side margin space would be mostly essentially wasted, which is why I experimented with putting the URLs in there. The Handbook too, probably being written or converted for the Web, has very few footnotes, but the Tufte layout with it's margin (great for writing on it) increases the page count, arguable if for good use. So if URLs shouldn't go into the Wrap side margin in Tufte layout (don't forget, Peeragogy layout is always available too), what should? Writing more footnotes for the sake of it, so the PDF print layout has more stuff distributed over the page (which doesn't translate to the EPUB or Website necessarily) wouldn't make much sense, would it? And then, such need to be supported by Markdown (not sure if they are, but maybe they are in the GitHub flavor of Markdown, which may in some form survive Pandoc as the Handbook v3 features at least one as it seems, which wasn't manually added to the .tex, which might be what was done for the workbook).
A regular newsletter can be a good vehicle to keep observers updated; to potentially engage them with calls-to-action, surveys, etc.; it might help to coordinate activity (as some data of or besides the Monthly Wrap might also go into the (live in terms of more or less "real-time"?) Dashboard) and along the lines of an Engelbartian Journal could be another resource in the Peeragogy library.
The EPUB, haven't inspected it closely for both the Handbook (which already exists and is produced for quite a while) or the Monthly Wrap yet (preferrably on a real e-ink device), but would assume that it has less problems, especially of typesetting nature as PDF fixed-paper-print-dimensionality encounters, so only some regular HTML/Web fiddling might still be needed here and there, review of the metadata, and "cover" page. I also started to experiment with adding Tufte CSS to EPUB, but don't get too excited as in EPUB, the main Tufte feature of the side margin makes no sense anyway, I'm also not opposed to the notion that a browser or e-book reader device should indeed be a user agent (acting on behalf and in the interest of the user, enabling + augmenting him/her on a custom, individual basis instead of patronizing or artificially imposing restrictions as today's Web browsers do) for applying ViewSpecs configured by the user (where the authors/publishers layout is only one option among many others, not even the default one), and I didn't hear much discussion about the licensing of the Tufte layouts + fonts – what that would mean for combined works and their sharability/distributability/remixability (except undoing/stripping the Tufte layout again, which renders the whole point mute of having it in the first place). Technically, I'm also hesitant to add CSS parsing to my EPUB packer for the only reason to extract @font-face path references to the local font files so I can redirect/rewrite them to their local locations in the EPUB package, but there likely can be a hack/workaround for it specifically for Wrap and Handbook. CSS should have been specified in XML (and still can be, but then again, a converter would be needed for all the CSS out there that's in its own domain-specific language syntax), reducing complexibility and increasing compatibility + ease of use a great deal.