And a while back MAL reported that the EuroPython Society uses a Condorcet method (Schulze) for their board elections. The service they use always displays the full preference matrix in its results, because that is the voter input to that method. It’s only part of the input to STAR (which also asks for cardinal scores, not just preferences).
Several Condorcet methods also deliver top-notch results. The knock against them is the extraordinary complexity of their schemes for breaking ties. People are happier with schemes they can understand without a doctorate in graph theory 
They may make a comeback, though. In 2021, Sara Wolk introduced a variant called “Ranked Robin”, with much simpler tie-breaking protocols. That’s pleasant enough for “just folks” that even the service we’re using now (whose DNA is averse to needless complexity) supports a version of that.
In my judgment it’s technically fine too, but on a human-factors basis I think people come to like STAR better: on Condorcet ballots, there’s no possibility to express to what degree you like one candidate more (or less) than another. Just unquantified preference order.
OTOH, if you have 100 candidates, a Condorcet ballot allows (or even requires, in some places) imposing a total preference order on all of them. STAR remains fixed at 6 levels regardless of how many candidates there are.
So if someone wants to push for Ranked Robin, that’s a fine choice to me too, but I won’t be the one to do it.