Maybe LLMs will solve this. Have them read the code and then be able to ask them questions about how it works?
It's not just docs though. Maybe it's going over an idea. "I'm thinking of solving this issue by doing X, what do you think?" Same, WFO, immediate answer. WFH, answer in 5 secs to 5hrs+ or never.
People will complain that getting a question takes them out of the zone. That might be true but it's never been true for any co-worker I've ever personally worked with. Nor with myself. It's always been easy and pleasant to answer a coworker's question. A few times a year I'm working on something so complicated I need to be uninterrupted for a few hours but that's rare, for me at least.
One of the funniest changes that came with this is that, even when everyone is working in the office now, we still communicate primarily through our WFH chat program because it is less intrusive.
It's also very handy for referring back to later — to the extent that when we do have an in-person discussion, we usually summarize it in the chat afterwards so that we can search for it in the future.
Mind you, we don't have the issues you describe (questions going unanswered for hours or not at all). With very few exceptions, such as when someone was out sick and I didn't realize, all of my questions have been answered within minutes. I imagine that is is because we are a relatively small team (albeit stretched across several timezones these days) and the chat is mostly pretty quiet.
My only complaint is that it has made the separation between work time and leisure time a lot more fuzzy. We all basically act as if we are on call all the time to answer questions, and it is common for my coworkers to even attend video call meetings from doctor's offices and overseas vacations just to say hi.
Unsure if you're attributing it to Feynman (although I do see the relationship in Feynman's thought process, but the quote is from Einstein:
> If you can't explain it simply, you don't understand it well enough.
... buuut really it's from Nicolas Boileau (1674):
> Ce que l'on conçoit bien s'énonce clairement. Et les mots pour le dire arrivent aisément.
Tx'd:
> What is well understood is told with clarity, and the words to say it come up easily.
As well as a few others that are right on point:
> Avant donc que d'écrire, apprenez à penser. Selon que notre idée est plus ou moins obscure, l'expression la suit, ou moins nette, ou plus pure.
Tx'd:
> Before even to think one should learn to write. Whether the idea is more or less obscure, expression follows through, 'ther less sharp or more pure.
In the office, quiet spaces to take a meeting from are scarce, but it's better established that your unscheduled time is actually spent on work.
By the way, LLMs really helped us here; we chat with a bot reading and extracting everything and storing it conveniently so it won't get lost for later.
On the flip side, open source projects function just fine with 100% remote work in different time zones.
One thing I found with WFH pre-and-post Covid is the the 'Feynman moment'- "If you cannot explain something in simple terms, you don't understand it". Complex architectures in the minds of a few people, or the way creaky systems worked together, and so on. Or to put it another way, no documentation for offline folks, because no one considered it important. So much for all that boasting about business continuity plans.
Which is almost a justification for being in the office, just to ask 'those people' how things works. It should also be a big red flag to management that things need fixing. But that's in the category of the management not seeing the financial benefit of doing it as there isn't an instant measurable up-front saving.
(edit) 'those people' are typical senior devs, and senior devs are often most likely to want to, or can, WFH.