With .gmi files or "gemini://" URLs and a compliant Gemini client, I don't need to even need to load the document beforehand to know if it intends to execute code on my device or not. It already won't by design, it won't in the future, and it doesn't require settings management, vendor whitelisting, popups, or caring who makes the browser for me to make it behave that way.
Whereas that .html document with it's noexec meta tag might be updated in the future to suddenly contain code.
With a dedicated Gemini client I simply have to trust/verify code provided the client developer.
With your solution now I have to trust/verify code provided by the browser developer(s), the apparatus the browser provides for extensions, and code provided by the extension developers.
If I'm super paranoid I can just look at a .gmi in Notepad or vi and understand it. I can't do that with all but the most basic HTML.
Ok I guess if you are that level of paranoid - even though both Chromium and Firefox are open source and under a heck of a lot of scrutiny for security vulnerabilities - then I understand why you prefer Gemini.
I just feel the fact that it cuts it self off from the wider clearnet completely kills your audience reach, if you’re ok writing to a very small insular community then sure, but most people want their writings to be read by as many people as possible.
This will let you create search engines that crawl and index these sites specifically.