It sounds like you're actually requesting a change to the parser's behavior, to merge consecutive lists into one even if they are separated by a blank line. Which might be a good idea.

James, do you understand the accessibility problem this creates?

This:

* One

* Two

* Three

is read by a screen reader in these words: "List of 1 items: (bullet) One, list end. List of 1 items: (bullet) Two, list end. List of 1 items: (bullet) Three, list end."

With correct formatting:

* One
* Two
* Three

the user only hears "List of 3 items: (bullet) One, (bullet) Two, (bullet) Three, list end."

That's a pretty big gain for these readers.

I've never seen a good use of multiple adjacent lists; have you? And, especially, I've never seen a good use of multiple adjacent lists that only has *one* blank line between the list items, which produces a rendered formatting difference that is barely visible.

It would also completely solve the problem of unintentionally splitting lists when you're trying to add a bullet in the middle of a list, which you can see at https://en.wikipedia.org/w/index.php?diff=690536710&oldid=690522879 and this screenshot.

The list looks correctly formatted in VisualEditor, but the wikitext produces three separate lists, and correcting it is a significant hassle, especially in nested lists.

Points well taken here but let us not lose sight of the fact the "blank lines" between list-items are only an issue under the wiki mark-up. In a "normal" html5 world, it doesn't matter how many so called "blank lines" exist between list-item elements (depicted below)....

<ul>
<li>Terrestrial planets</li>

<li>Other planets</li>


<li>Giant planets
<ul>
<li>Gas giants</li>
<li>Ice giants</li>
</ul>
</li>
</ul>

... the top-level list of planet types will render as a 'single list' regardless. This difference in "behavior" I'm guessing is because a "blank line" is not considered a 'new-line-start' normally as it seems to be the case under the wiki mark up.

In other terms, this...

* One

* Two

* Three

... is, at some point, 'on the way' to final rendering (during the wiki-markup processing), 'seen' as...

* One
<p></p>
* Two
<p></p>
* Three

... and that's why a single list of three items is broken up into three separate lists with one item each. Those empty Paragraph tags get stripped by the end of the process however; they just happen to exist at the wrong time and in the worst place in the current approach is all.