I'm creating a couple of HTML5 games that have multiple parts or different skinned versions of the same game.

I'd like to be able to list each part under the banner of a single page, with the ability to label a button or configure a UI element per part. Not only that, but I want to create variations of the same game.

Nothing too crazy, but maybe a maximum of 10 entry points per page. Currently, web based applications only allow one clickable entry point, and I'm wary about spamming Itch with an individual page for each variant. It smells like a SEO tactic and would simply annoy users of the site. I don't think that users would via the content of a single game page having more variants in the same negative light.

Of course, I can include multiple versions of the same game / web application, if zipped up, as downloads, but I really want to be able to be able to click to start.

Possible Implementations:

1- Allow a "Hero" web app, which is part 1 (or the default version). Next, let me "add" additional apps, specifying the zipfile / index.html containing each part, and also being able to label the buttons for additional parts. Maybe a tiny thumbnail for additional parts, but that's not as important as the customizable button lable. When clicking the button, it should load the other part, and go into fullscreen mode.

2 - Another possible implementation is to provide a dropdown or radio buttons of the different parts, with a default being able to be set. The site would remember the last one selected for return visits to the page. When selecting a new option, the iframe updates for data relevant to that part (the button text should make it clear which variant will run when clicked).

You may suggest work arounds inside my own web app, such as an in built menu system, but there is also benefit to this being external to the games themselves, not least being able to silo complexity.

Just an idea.

I get what you are saying but I also feel that the design/implementation is on your shoulders rather than that of the itch.io website.

The primary focus should be on end-user experience first and foremost.  Multiple-'click to play' things that aren't YT videos are a negative end-user experience.  Even when they are YT videos it can force the page to load much slower than normal.

Consider using a game engine (like GoDoT) that natively supports exporting to HTML5 so you can fully experiment with your ideas standalone prior to making it a 'web app'.

maybe you could create a single html5 project that has a launcher that lets people select the project they want to run? essentially once someone starts your iframe, you can show them whatever you want.