I'm not convinced that Heise is any better than Petrus for FMC purposes, assuming that these methods are used in conjunction with NISS and insertions.
Fixing the pseudoblocks alongside solving EO has a move count cost, and ultimately is very similar to how you can often influence EO in Petrus solves by adding 1-2 moves, or failing that, switching to inverse then adding 1-2 moves. I guess if you somehow "forbid" using NISS, then Heise is a better version of Petrus simply due to having the option to build pseudoblocks, but that seems like a silly restriction.
Of all the methods listed in the poll, Petrus has been giving me the best mileage (majority of my FMC solves somewhat resemble Petrus; most of the remainder are EO first) so that got my vote, but domino reduction is predominant among top-level FMC solvers.