I don'y understand your attitude. I am not a yacht designer and I never have pretended to be. That does not make less true that today all major NA cabinets use CFD in what regards designing hulls, keels or rudders. Many advertise so on their web sites and others even publish some computer prints of the studies. It is only a complex computer program that accesses the efficiency of what has been drawn and contributes to improve the design.
Paulo, I am not sure that your assumptions are right. I attend a lot of presentations on the development of yacht design software and how it is being used. Clearly CFD is out there and being used on specfic projects by the high visibility high performance shops. But for the most part, what seems to be way more common in design offices are a battery of VPP's (velocity prediction programs) that are based on historical data and prior CFD and towing tank studies. These VPP's are pretty good but they do not analyze down to a level of detail which would validate the performance of a particular choice of keel profile, keel section, or keel layout.
The better of these programs produce quasi CFD data and the kinds of graphics that replicate what you would expect out of a full blown CFD analysis, but these programs do not appear to be at the fine grain level that would be needed to analyze whether a specific tandem keel worked effectively or not.
This is not to say that RANS and CFD analysis is never performed. I think that really high budget projects get that level of attention. Farr's office showed amazing graphics from their analysis of the new Volvo boats.
I am talking with Quantum about doing a paper exploring the trade off between sail area and sail shaping for cruising boats (i.e. more powerful sections vs. larger sail area) and in the course of the preliminary conversation, it was suggested that ideally we should use the VPP for a specific design, including the drag, leeway, and stability characteristics of a real hull and rig.
The designer at Quantum and I spoke to one of the partners at Farr about the likelihood of getting basic data on one of thier newer production designs, and while it was clear that Farr ran moderately sophisticated performance analysis on all of their designs, it was not clear that it was run on a fine grained enough level to allow a precise enough computer simulation for the sail study.
As recently as only two years ago, there was a paper presented on a study to validate a CFD analysis of flows around a keel. It was one of the first times that they were able to instriment a full sized keel in action to measure the points at which laminar flow became turbulent. They found that there was almost no steady state laminar flow, and that the flow even in calm water was discontinuous, shredding in much the same way that spinnakers have discontinuous flow. It largely invalidated the CFD which assumed way more continuous flow than existed, and it was only once factors were added for discountinuity that the CFD could be validated.
So, while I do not want to talk for Bob, and I do not want to be dimissive of what you are saying, my sense of this is that CFD, especially for keels is still in its infancy, and that while better design firms have access to CFD for their higher budget projects, it is not in general use and perhaps fine level analysis not accessible enough to acertain that a tandem keel really does work as advertised.