Condorcet voting systems are normally posed as a solution for
single-winner social choice problems.
Simple multiple-winner Condorect (ranking the candidates, and taking the top N as winners) is subject to some particularly nasty pathologies, which I don't much want to get into.
Nor do I much want to cover this topic in a diary ... but those who feel impelled to discuss, go knock yourselves out. (In case of multiple survivors, feel free to proceed in multiple-knockout tournaments of your own design.)
By most measures, IRV-2 would tend to produce maximum satisfaction. Condorcet-2 would tend to leave large factions unrepresented.
Popularity in a low-turnout election is a weak indicator of merit. (Intended here as a comment on merit of voting systems, not merit of Senate candidates.)