Good point. I would be fine with any of the above-mentioned, as long as voters were voting for people and not parties (as is often the case with PR).

Parties are fictions, while people are real, and classical liberals have always resisted enfranchising factions.

Another concern with RRV and Asset Voting is that they seem more complicated than plain RV. I think common understanding of the voting system is essential to ensuring all voters feel the process is fair.

That said, I noticed on rangevoting.org an off-hand remark that the best choice is selection by lot. I whole-heartedly agree with this statement. It’s easy, conflict-free, and endorsed by the Greek city-states.

However, I found it interesting that the endorsement seemed to conflict with another part of the site which listed sortition as having the highest Bayesian Regret. Perhaps I misread the data? Or maybe that only applies to single-winner elections?

Whatever the case may be, range, reweighted range, asset, and sortition would all be a huge improvement over the first-past-the-post model now predominant in the Anglosphere.

http://rangevoting.org/RRVj.html

Asset Voting is also nice, and much simpler:
http://rangevoting.org/Asset.html