My heaviest lens is a 70-300L, so the 'moderate' duty heads should be the right balance.
(...), having an off-center load increases the chances of vibration. (...)
Sorry for being pedantic but I read this very often. Actually, having a very balanced setup *increases* the risk of vibrations as the system can move more easily around its support and is thus more prone to pick up any high frequency vibrations. The greater the mass or the more off-centred the mass is placed, the more difficult it is to move the system out of its (in-) balanced pivot point. It is best shown by the fact that a well-balanced gimbal supported system (like a heavy telephoto lens) can be moved easily by a push of a single finger whereas an off-balanced system requires more effort and force. That's why it is good to rest your hand firmly on the camera (body) while shooting from a tripod or by even placing a bean bag on the camera. This not only increases the mass of the system which then adopts a lower resonant frequency but it also helps to imbalance the sytem and stabilises it further.
Interesting hypothesis that an unbalanced load is less susceptible to vibration, and easily invalidated by just such a test as brad-man describes.
Partly for kicks, but partly to get a feel for the magnitude of the effects, I tried a few tests...
First off, the with the Canon tripod collar and the appropriate RRS plate, the 1D X + 70-300L can't be balanced – it's back-heavy even with the plate slid as far forward in an RRS clamp as the safety stop in the plate allows. I was able to balance it by loosening the plate and sliding it back (not ideal for regular use, as it defeats the anti-twist. I suspect a non-gripped body would balance properly. Even with the plate installed correctly, the load was less unbalanced than connecting to the body plate. I did the testing with the plate slid back for a balanced load.
In all cases, the setup was on the first floor of my house on a hardwood floor. I initially tried FoCal testing there, but the floor transmits enough vibration that I moved my AF testing to the basement. Still, the point here was to see vibration, so I chose the hardwood floor. I used Live View at 10x magnification, with the 70-300L at 300mm, focused on the far wall, and initiated vibration in a roughly reproducible fashion by dropping a sandbag from a fixed height to the floor. That resulted in vibration that was stronger than my normal shutter press for the same setup.
I first used my RRS TVC-33 and BH-55 LR, a very robust setup. Honestly, there was no easily discernible difference between the balanced setup on the collar and the unbalanced setup on the body. In both cases, the vibration damped in just over one second.
I next tried a less robust setup, my RRS TQC-14 and BH-30 LR. 'Less robust' is relative – the TQC-14 is RRS' lowest capacity tripod, rated for 25 lbs., but many will recall the picture I've posted of RRS owner Joe hanging his full body weight from it (body as in corpus, not camera). With that setup, there was a barely discernible difference between the balanced and unbalanced loads, with the unbalanced load taking slightly longer to settle. Still, in both cases the vibrations damped in around two seconds.
To accentuate the effect, I raised the center column of the TQC-14 (something I don't normally do). In that test, the damping times were long enough to time with a stopwatch. The balanced load settled in just under 4 seconds on average, and the unbalanced load settled in just over 5 seconds on average. So, a bit over 25% longer to damp the vibrations of the unbalanced load.
Finally, I switched to the 1D X + 100-400L at 400mm (that combo could be balanced properly on the tripod foot with the lens plate in the correct position). The balanced load took only slightly longer to settle than the lighter/shorter FL 70-300L, damping after just over 4 seconds on average, while the unbalanced load took just over 6 seconds to settle. So, at 400mm the unbalanced load took about 33% longer to damp.
A couple other random points: I didn't try a beanbag on the camera, but placing my hand on the camera was a bust, my own random physical motion prevented the image from settling at all, as long as I was touching the camera. Hanging a weight from the hook under the tripod platform/center column shortened the damping time, as one would expect.
While I wasn't surprised that the unbalanced load was more susceptible to vibration, I was surprised that the effect was relatively minor. Even in basically a worst-case scenario – 400mm lens on a raised center column of a travel tripod (albeit a very good one), a 10 s self-timer would have been sufficient (of course, that assumes the source of vibration isn't continuous).