Previews are different from native apps though, and when you click on a link to a document, you only have to click "Open" to open a native app. How is this an issue? For quick edits I can see the appeal, however.
> Installing a separate app may not be possible for aging parents, people using shared computers (e.g., library or school computer), or organizations with restricted IT systems
This seems like a non-issue, too. If my grandma were to use the PC, I would be the one installing programs (or the OS) on it, and if I knew she wanted to open and edit documents, I would install LibreOffice. It is one of the programs I install on my parent's computer. You (or the IT staff) ensure tools are available; you install LibreOffice, similarly to how you would install a browser to begin with. For example IT staff are already responsible for installing and maintaining browsers. Adding productivity software to their responsibilities is a standard practice in professional environments.
Browser-based tools are a complement, not a replacement, and the push toward browser dependence should consider the trade-offs in performance, user experience and so forth. Using browsers for productivity tools may feel more convenient in very specific edge cases, but for most users and workplaces, the effort involved in installing and using native applications is negligible in comparison.
(I remember the old days when I used https://gobby.github.io with my friend to edit code together. It is a native program for editing documents (incl. code) together. Today you would probably use an extension to VS Code unless it supports it natively (no clue).)
I am pretty sure there are people on here who would be able to explain it better than I do.