There are also some accessibility concerns with the current implementation IIRC where text (via content on the pseudo element) on the active track is being read even when the switch is off and vice versa.

Putting text on the thumb or track is a bit of a hassle, but then again that might be too much of an edge case. Some Design Systems use a button as a base for the switch, but <button type="switch"> would likely be less useful because immediate action can be scripted for the checkbox.

It's unnecessarily hard to figure out if an element currently can be focused, respectively participates in the sequential focus navigation or not. `element.focusable` would do the trick, but right now, handstands are necessary to appropriate that functionality.

I had to come up with a custom solution to write an experimental polyfill for the focusgroup proposal (github.com/gfellerph/fo...). It's simplistic and uses basically the API that Lea proposed. Lifecycle hooks are being called when elements enter or leave the DOM and it can be scoped to a subtree.

I take part in these surveys because I'm very interested in the results. They are relevant to my work as a dev to see hypes and get early warnings for fading frameworks and I learn a lot. Making it longer to cover more ground would be a benefit for me (an almost perfectly average respondent).

It's the missing piece together with interest invokers for a native html+css floating-ui tooltip replacement. I guess it would even be possible to draw a safe space for cursor movement between the tooltip and the element with this in place.