Almost nobody uses UFS in FreeNAS because the purpose of FreeNAS is to manage network storage, and usually the people smart enough to want this are looking for the features. But some of us have done FFS. It fits certain use models.
Almost nobody uses ZFS in FreeBSD because it's still regarded as somewhat experimental, slow, and piggy. I think they finally worked around some of the "really annoying to install" issues.
Now, 99% of FreeBSD-derived installs are FreeBSD, not FreeNAS, so from my point of view, the FFS/UFS installed base is effin' gigantic compared to ZFS. The FFS in FreeNAS is no different than the FFS in FreeBSD, so for the purposes of this discussion, what I see are millions of filesystems that've successfully stored files over the last two decades, while ZFS's track record is more ... um, recent, is that a polite term? (did someone actually suggest ZFS could cause a kernel panic just a few messages back?)
The point is that it's probably a little unfair to place such emphasis on data reliability features. It is kind of like me saying my vehicle is safer in a car crash. If that's because I'm running around in a Hummer, it fails to recognize that I'm probably going to kill the other guy, and that I'm making a much larger environmental impact with pollution. For the most part, people losing bits in a way that ZFS would catch but UFS would not is a very unusual event. On the flip side, ZFS tends to make storage systems more complex and therefore more breakable. Sure, you can blame the user, and that might even be fair in many cases, but lost bits are lost bits... there's a little bit of irony in a system with such great data reliability features being able to lose bits more easily.