Desktop development in the past was locked by widget based GUIs and their limits, they are not classic editable DocUIs, modern WebUIs are the commercially needed answer since they are DocUIs BUT not under user control. Most modern devs simply use WebUIs, even locally, because they can't even imaging an editable DocUI. When some colleagues see rare presentations from me (I talk a lot, but hate slides) in Emacs/EXWM/org-mode/dslide can't even understand what their see. Live rendering of an active markup witch is indeed free form text eventually with executable code (elisp: links or C-x C-e full displayed sexps) is something they can barely understand, having the same language, functions available everywhere is totally alien for most.
Even if we have the code, if the coded software is something unsuitable for personal, SME usage being just a copy of some big tech walled garden, the code alone does not suffice. Having the entire Alphabet infra under AGPL does not fit much any FLOSS purpose. It just allow big tech interoperability witch is their biggest Achilles's heel since having no walled gardens, an OS as a single application in user programmable hands interconnected with other desktops as well, is the strongest FLOSS point no one company can aspire to match no matter their size.
FLOSS need a FLOSS IT paradigm BEFORE the code, we have more than enough code not to care about "OSS Enterprise" models, we miss instead the knowledgeable user base and desktop paradigm.