19? Less than me, then. Anyhow, first moving the goalposts, then switching to ad hominem attacks. No wonder you aren't convincing anyone.

BTW, I had a quick look through your posts. I see you are as misguided on LVT. And don't get me started on how clueless you are about Functional Programming. Bye

3. Surely you know that no system satisfies all criteria, including Approval. Schulze is actively used, proven, and relatively straightforward if you know graphs.

2. So, an implementation of sortition. Not my exact preference, but not miles away.

1. You can model them as either, and they can be useful models, but it's important to remember that Models ≠ Reality. In real life, voters have complex per-issue preferences. Electing representatives destroys most of this information, but PR supports per-issue coalitions better.

The Spanish system uses provinces as districts. This works well in the populous coastal regions and Madrid but fails horribly in the Spanish hinterland.

The UK elections would have been roughly the same with approval as plurality because of tactical voting.

The point that you seem to be missing is that PR systems support and work best with single districts or approximations of it (e.g., Germany). Gerrymandering is not possible at all.

Now you're just moving the goalposts. Sorry, but too many threads already open.