What appears to be lacking (the amateurish product information page is not a big help) is an integral rotator for the yaw adjustment. Which means you'd have to line up the markings on your regular (ball?) head or add something from another manufacturer. No idea what the lens retaining strap is for; the lens mount seems to keep my lenses attached on its own. Completely pointless, I'd say.
Personally I'd go for something like a base model Nodal Ninja. I'm still using an early model NN3 for VRs with either a D200 or D700. They are a bit too heavy but it just about works. We're not talking here about multi-row panos using large heavy lenses, however I have done them experimentally using a 14-24 2.8 - which is barely feasible on this rig.
If you're primarily going to do VR panos they will end up displaying on a web browser. Most of the real estate applications require small .swf files, typically <2Mb, so preparing huge files is a waste of time - IMHO. There are obviously exceptions to this, but they aren't a common requirement. Most people doing VR panos use fisheye lenses in the 10-16mm range, depending on whether you're using a crop or full size sensor.
Most people on crop sensors are using the 10.5 full frame Nikon or 8mm Sigma circular lenses. I've tried using the 10.5 with the mini lens hood shaved off as a circular fisheye on the D700 but I'm not convinced there's any real advantage over shooting with the D200 (IQ issues aside). I have a Russian Zenitar 15mm lens too which works reasonably on the D700 @ F8 (which is about what's needed for VR panos in constrained interiors) however it's very susceptible to flare.
On the number of shots issue, assuming a 10.5 mm lens on DX or 15/16mm on FX you'll need to shoot six shots round @ 60 degree yaw intervals plus one zenith shot (up) @ +90 deg tilt. This can be circumvented by shooting the equatorial set at a slight +ve tilt, but why bother? It just compounds the nadir shot problem. For nadirs, using a pano head like a Nodal Ninja (or one of several equivalents) which has a small footprint, requires 2 shots @-90 deg (down) each at 180 deg of yaw separation. Alpha channel masking of the head component in these shots, when the set is stitched, will leave a small central hole which you can fill using a variety of techniques including the "mirror ball" effect or corporate logos etc.
For a properly patched nadir hole you'll need to use the "viewpoint correction method" ( offsetting the tripod and shooting at about -60 deg toward the place where the tripod formerly sat) or a hand held shot (not recommended) or one of a variety of odd methods derived by various VR pano gurus. But this is a refinement. So, HDR exposure bracketing aside, at these focal lengths we're talking about a maximum number of shots @ 10, or 9 if the nadir patch is omitted. This could be further reduced to 4-round (yaw), possibly tilted up, to eliminate the zenith, with two nadir shots and a bodged hole. Which would amount to a total of six shots. Given that it doesn't take long, once set up, to shoot either six or ten shots, the savings in shooting time are insignificant. The increased load is in the pp - particularly if you're shooting raw and HDR.
I use PTGuiPro but many also swear by Autopano Pro ("Pro"... groan.) Assuming a correct setup for the entrance pupil (aka no-parallax point) and a little practice, the stitcher makes short work of the job on a fast PC with plenty of ram. The time consuming stuff is the raw conversion, alpha channel masking etc. Of course if you've got the work you can batch process all this stuff - but if you've got the work you'll buy an accurate pano head that enables you to do reproduce-able template stitching too.
Anyway that's the short form derived from my own experience - I've done a couple of hundred VR HDR panos in all kinds of human-free environments. Populated environments is another subject.
Roy