A midships wand is definitely better than a bow wand in flat water because it can be significantly lighter if the Bradfield system is followed.
In chop with a forward wand there is a delay between wand activation and the foil being where the wand was when activated of around a quarter second at upwind speeds in addition to wand response(bungy) lag. This intereferes with the ability of the forward wand to average it's inputs.
The midship wand has only bungy lag and therefore has a quicker and more reliable response to begining to average wand input resulting in a steady altitude and less of a likelyhood of dramatic flyouts from waves. This can be seen on the Rave video in the rapid movement of the wand which is noticeably more than on a Moth in similar conditions as best I can tell.
This is very important: Because the forward wand's center of lift(drag) is significantly displaced from the boats center of lift when flying there is another element added to the forward wand response and that is pitch coupling. This can add or subtract from the wands response-when it adds the boat flys out of a wave. That happens much less frequently with fewer consequences on a boat equipped with a midship wand where there is NO PITCH COUPLING whatsoever.
So the combination of the midship wand developing an averaging response quicker than does a forward wand and the pitch coupling common only with a forward wand explains very well the observed responses of both boats and seems to suggest that a well designed planing wand mounted midships might be an improvement on a monofoiler in every conditon.
At 22 knots there is no 'anticipatory' or 'reactionary' wand-just an averaging system with fairly rapid wand movement on a Bradfield system.
The bow wand would be subject to unanticipated moves due to pitch coupling as mentioned earlier. I'd bet that the midship wand would be -by far-best in these conditions.
The question about foil spray is one of the best yet. On the Bradfield system the wand is next to- not in line with -the foil and is not affected by the foil "spray" in any way. But in applying the Bradfield system to a bi-foiler the wand will need to be close to the center line and the foil wake and/or spray may or may not have some effect.
The Japanese Mothies are experimenting with an in line trailing wand system similar to the ones used for many years in human powered foilers(very successfuly). That type of wand does not have the power of a Bradfield type and yet has worked very well despite being located behind the vertical fin.I'd imagine that the Bradfield type would work well-I'll know before too long on my boat. It would be interesting to see three equal Moths rigged with the standard ,Bradfield type and Japanese type.
The problem with the bow wand seems to be that it can produce movements of the flap unrelated to the averaging of the wand.Depending on whether it adds to or subtracts from flap movement the average wand movement can result in unpredictable and sudden big movements of the flap.(Probably exacerbated with flexible wands)
This kind of pitch coupling can't occur in the Bradfield system because the wand makes contact with the water at the center of lift of the boat so there is no pitch coupling at all.
Unless the wavelength is so great that the boat contours the averaging occurs when the boat platforms and results in rapid wand movement(several times per second) while the boats altitude remains the same. In wave conditions where the boat is platforming the wand reacts multiple times per second and the forward wand anticipates nothing-in fact there is some reason to believe that in those conditions the separation between the center of lift of the wand and center of lift of the boat is not a positive thing at all-contributing to erratic behaviour. Flexible wands would tend to agravate the problem-both in averaging and with pitch coupling. Depending on whether the wand is a planing wand or not, conditions can make the ride feel like a bumpy road in short steep chop.
One thing all this recent wand study has done is reinforce 100 times over the fact that a manual system that mixes main foil control with rudder foil control will be superior in racing. Wand movement at multiple times per second reduced to virtually no movement! When the wand is averaging it is essentilly creating major changes in the lift coeficient of the foil rapidly and for no net change in altitude. This is a drag producing machine!
Manual control will be fairly hard to learn to race well but the potential savings are much greater that I thought just a few weeks ago.