The chief indicators of noise are blade speed and furling mechanism. Turbines with high tip-speed ratios (>5 or 6) are loud due to tip spillage and turbulence. Many small, light turbines (like the SWWP Air models, TSR=9) use slender, high-speed blades to compensate for their minimal electrics -- the "small alternator spun fast" design school. There's all kinds of problems that go with that. Sure, they spool up to charging voltage quickly, but the coils saturate at low wind speeds, they quickly outrun their magnetic flux, output flattens, and then it's just noise noise noise until something flies apart, which doesn't take long usually.
To protect gracile coils from sustained high winds, many small turbines employ some kind of furling or overspeed protection. The five most common methods are:
The Fly Ball (pioneered by Jacobs and most common in larger turbines), in which governing weights slide out from the hub and increase angular momentum; I THINK the KISS units employ something like, Cam will have to comment on that. This method is quiet but has limited governing range.
Blade pitching, where the attack angle is reduced until the blades 'luff' and depower. This is absolutely the BEST means for preventing overspeed and for keeping a turbine generating at peak regardless of wind speed. Alas, it's hard to fit the mechanism into the hub of a small turbine. I don't know of any marine turbine that uses variable blade pitch, but maybe one exists. It is quiet.
Aeroelastic blades: Do Not Go There. In which the turbine governs speed by allowing its blades to flex and spill wind. Good on paper; in reality, the vibrations and flutter set up by independently actuating blades is hell on roller skates. Early SWWP Air models used aeroelastic, and they were terribly noisy. Proven Wind Turbines are downwind models that use an 'oragami' hinged blade that furls the blades conically; clever theory, subject to stresses, don't think there's a boat-sized model.
Side furling: Turbine's yaw shaft is offset from the centerline and the tail is on a sloped hinge, so at a set windspeed the blade plane turns sideways to the wind. This is an elegant and dependable method of overspeed protection. It can also be fairly loud, since you have some blades going upwind and others downwind. The differential loading is also hard on bearings and blade roots. My Bergey works this way, but it's on the hard; I suspect a folding armature might cause trouble in a rolling sailboat.
Finally, there's alternator braking. The leads are shorted together, often many times a second, placing maximum drag on the magnet rotor. Sounds like a fabulous idea, and many small turbines use it. Problem is, a small turbine at governing speeds already has its coils close to saturated. Sustained high winds may cause the blades to 'break away', overcoming the breaking force. Then it's mayhem. Stators melt, rectifiers catch fire, wires burn off, and the blades can accelerate to open-circuit levels. That's REALLY noisy! I can always tell if my turbine has kicked its circuit breaker because it is incredibly loud. Even when its functioning properly, the pulse-loading of an alternator brake can make a queer growling or strobing sound.
The engineering binds inherent in wind turbines are brutal; those inherent in small wind turbines are insuperable. They capture little wind to start, and to get any useful amperage out of them, they require high rotational speeds; to keep them light, they use small electrics and run at the edge of melting. Noise is almost inescapable, given those realities. I'd look for a low-speed unit with a tip speed ratio under six, big rare-earth magnets and coils, good MPPT circuitry to boost charging voltage, and a governing system you can live with. If you want silence, PV is the way to go.