Fix sibling inserter hit area

[jasmussen]

This fixes #8881.

The sibling inserter hit area was not centered perfectly between two blocks. This PR fixes that.

Props to @chrisvanpatten for this fix.

The issue, as described in #8881, is that the sibling inserter covered actual text of text blocks. Specifically, it has a large empty hit area whose sole purpose is to show you the sibling inserter button when you hover that area. However this was misaligned, and covered text below.

This fixes that by aligning it correctly. Here's a GIF with debug colors. The orange shows the sibling inserter hit area, the red shows the text input fields:

Noting to myself that if we move further with #8350, we need to account for that with the sibling inserter positioning and height. I have some ideas, including simply having a smaller minimum height of the hoverable area, but this is something to think about. The code in question is this:

[chrisvanpatten]

Is there any reason not to just turn this off on selected blocks entirely since we don't show an inserter there anyway?

.editor-block-list__block.is-selected > .editor-block-list__insertion-point {
  display: none;
}

The only case this could have an impact in the future is if we decide to show the insertion point on selected blocks without toolbars. It'd have to be rewritten with a selector like .editor-block-list__block.is-selected:not(.has-toolbar). But that's a concern for the future!

[jasmussen]

Is there any reason not to just turn this off on selected blocks entirely since we don't show an inserter there anyway?

Yes, you have to be able to tab into it. It's not super ideal, and something that can be looked at separately as a future thing. But it's a rather biggish refactor to change this with tab order and such.

[chrisvanpatten]

TIL the inserter is available via tab even when hidden from the mouse!

[ZebulanStanphill]

Speaking of which, I really think the sibling inserter belongs below blocks, rather than above. It makes so much more sense, in my opinion.

[noisysocks]

Looks good. Not noticing any regressions 🙂👍