I'm not sure about the floating element distinction on the RB lenses, but the RZ lenses with floating elements carry the "L" designation, and instead of having a DOF calculator that spins 360 degrees to show imperial or metric scales, it has just the floating element ring, which does double duty. Anyway, that has all been noted here already, but I wanted to offer that my experience with the non-floating element 65 when I used one owned by work, and the later L-A version with the floating element has been that proper use of the floating element makes a noticeable difference in sharpness and reduced image distortion across the field, and has an especially noticeable effect in field flatness towards the corners when shallow DOF is in play. The premium is not that high on the used market, and is well worth it.
The last thing to consider is that the price we pay by having to manually set the floating element is more than made up by the fact that our bellows focused lenses do not suffer from the design compromise of being forced to optimize a lens for near or far focus at the drawing board. I remember this discussion several decades ago at a large format workshop that I attended, but it slipped into history over the years until I was reminded of the issue on the Hasselblad site a few days ago.
Hasselblad wrote a detailed paper on the reason they dropped Zeiss as a lens making partner, and went with Fuji for their H series lenses. They honored their long relationship with Zeiss, but explained that Zeiss made lenses for Hasselblad without allowing much in the way of direct development involvement from Hasselblad engineers. The result was that Zeiss lenses were provided with Zeiss built shutters, and designs were optimized at infinity focus. That was simply how Zeiss did things, take it or leave it (that's a paraphrase).
Fuji, on the other hand, was eager to do the team design thing, and had the computer software to allow the teams to be half a world apart. The end result is that the new H series lenses sport Hasselblad designed and built electronic shutters, and Fuji glass and barrels that are assembled at Fuji around complete Hasselblad shutter/iris assemblies shipped from Sweden. Pretty slick, no? It gets better, and this is the forgotten point....
The new Hasselblad lenses are also helical focused lenses as in the past. This is a big space saver, as the focus is made by moving lens elements, and does not require as much motion as with bellow systems (how well us RB/RZ users know that). The downside is that helical focus lenses cannot be completely corrected for aberrations at all focus settings, as lens element groups are moved to to achieve focus, and that changes the corrections needed. So, a choice must be made. Near focus correction, or infinity focus correction? Zeiss chose infinity focus optimization, and according to the charts posted by Hasselblad, that gave stellar performance at infinity, but lesser, and sometimes downright low grade performance at closer focus ranges.
Their thought was that most people using their gear are shooting closer than at infinity, as they are shooting portraits, studio, product, interiors, etc. When they partnered with Fuji, they optimized the H series lens designs for close work, did the best they could do within reason for infinity, and corrected what errors remained at infinity with digital software. Obviously, it was a huge success. What surprised me most though, was how those awesome Zeiss lenses that set the standard at infinity really suffered at closer ranges.
That reminded me that our "old school" bellows-focused lenses do not suffer anywhere as much from having to choose near or infinity optimization at design time, as the main elements are fixed, and do not introduce additional errors to the mix as they are focused. If the lens is a well corrected design, it remains that way (for the most part) across the board. The floating element corrects for flatness of field errors, and had to be moved to accommodate the change in the angular nature of the light rays making it to the film or sensor plane as it's point of focus changes from infinity to close up. It's a much easier problem to design a single moving corrective element that isn't even critical in terms of positioning than to design a complex cam arrangement to move several elements in critical step though, and still have to accept compromise in the end.
That's not to say that our bellows lenses are perfect by any stretch, but the odd little routine we go through to focus our subject does have it's benefits....