It's a clever workaround that the state constitution says you can elect something with RCV, but they could run a preliminary vote that way. I almost wish they'd do some kind of 2-winner STV into a top-two runoff. (Which is probably a suboptimal election algorithm for social utility maximizing, and needlessly wasteful for getting people out to vote twice when once would do, but might have some odd realpolitik advantages?)
I just hope they release full data because I want more public datasets about how people actually use these kinds of ballots.