"Don't get me started" (that's a common english phrase that may not translate well. It means that you will provoke much heated debate, just for asking!)
I've often said that too much importance is put upon selecting the perfect airfoil, but for small backyard projects such as ours the result is hard to detect. Using the Clark Y or the NACA is perfectly fine. These are tried-and-true airfoils, easy to make, and have thick cross-sections for strength. You make your life complicated with the cambered airfoils, unless you are also ready to commit to making them in a somewhat complicated way (though if you do have a table saw or a router table available, then don't let this stop you.
In your case, Frans, you have electronics that monitor the turbine, and a pitch control system that responds to wind speed, so you are in a much better position than usual to measure and check these assumptions.
I believe the effort of making complex blade airfoils such as the SERI types is a total waste on a home-built wind turbine. Second, the only airfoil types to seriously consider are the ones designed to be effective at low Reynold's numbers (that aerodynamic terminology for "small scale"). Other than that, you have other things to do that are more important than the airfoil shape.
I usually see, in photos on this site, variations in profile that can add up to several millimeters, or in other words many % of the airfoil chord. In comparison, the difference between the Clary Y and NACA 4412 is also just a few % of the chord, in scale. To me, that means that the airfoils often used are not highly accurate reproductions of the "perfect" shape. The builder may have good things to say about the blades, but the airfoil isn't as true as he thinks.
Another fact is that the aerodynamic smoothness of the airfoil, which you should be sanding and finishing with paint or other protection, will be the strongest factor in the air drag. Since drag steals torque from the blades, any effort at reducing drag will pay handsomely in efficiency. Since sanding and painting should take about an hour or two of work, I would venture that the value of that time spent is worth 10 times the value of selecting a "perfect" airfoil. That task can take days, and the benefit will be smaller.
Those are just my general opinions about what's important to me, when making blades. Back to your problem:
Have you tried adjusting the pitch angle?
Is it the 12-foot or 10-foot blades that are giving you this trouble? Stalling always makes noise.
Is the "stall-effect" you mention an abrupt, cyclical one, or does it just rob power away? Shouldn't the blades be feathering at that wind speed? Have you compared the pitch of the blade to the inflow angle of the wind to find an approximate angle of attack?