1. The purpose of varying the pitch of a blade on a mill is to reduce the power pulled from the wind (and the drag from it) when the wind is strong.
That's one reason why variable-pitch turbines exist (pitch for speed control). Another reason why variable-pitch turbines exist is that the pitch angle can be optimized for the best power extraction when the wind speed is at a safe level.
I'll buy that.
(My discussion, though, was related to picking the shape of the blades, so I guess I oversimplified by ignoring the unrelated purposes.)
3. The profile of the blade is very low-friction, to avoid losses when spinning rapidly cross-wind. So feathering it into the wind shouldn't produce excessive drag.
I'm not sure what you're referring to here; if you're talking about controlling the speed of a turbine in very high winds, then excessive drag (to prevent the rotor rotation from accelerating), as well as a reduction in lift, is exactly what you want.
No. I'm saying that when the blades go more edge-on to the wind they're "slippery" rather than "draggy".
You don't need drag to slow down when you feather. Rotating the blades changed the tip speed ratio. (If the blades are turning too fast they'd pump air rather than being pumped by it (or if feathered enough would just swat it around) and THAT would produce the necessary drag to slow 'em down.)
4. Therefore, the appropriate design is to optimize the blade for the lowest wind speed, feather it (rotate the leading edge windward) by various amounts to tune it for higher wind speeds or lower loads, and not sweat the non-optimum shape when feathered.
If, by "lowest wind speed", you mean the typical wind speed for your site, I'd agree with this.
I think we're on the same page. What I meant was somewhere on the low side of the range of typical speeds. And size your prop so the genny will be maxed out for most typical speeds.
The point of a feathering prop is to be able to run the genny at its own max-out in most wind speeds. Then you pick your optimized speed to get the best you can from the distribution of lower wind speeds where you can't max the genny - taking into account that optimizing it for the wrong speed may slightly raise the point where the genny maxes.
What constitutes the "best you can" may depend on your situation. If slow winds are rare and only last a short time, it's OK to let the mill be pokey or even stop, and live on the stored power. But if low winds, even if rare, tend to last for days (and you don't have other resources except maybe an expensive-to-run backup generator) you may want to make the mill oversize and/or optimize the blads for lower winds, to keep alive during these times.
5. Non-optimum twist means the wind will produce some bending force in the blade (due to the hub and the tip wanting to turn at different rates).
There are multiple reasons to optimize the twist distribution. One is for structural concerns, which is what you mention here. The reason I refer to is minimum-induced loss, or greater aerodynamic efficiency. Optimizing for aerodynamic efficiency will not necessarily optimize for structural strength, though the two are not mutually exclusive (as far as I know right now).
I mentioned that because it was the only one I thought of at the time that I considered to be a concern. (Another is that inefficiency results in extra drag forces which must be fought by the supporting structure, and this may be an issue.)
I'd follow along some of the lines you give: optimize the rotor for the wind speed you are most likely to see. At low wind speeds, being close to feather gives high startup torque (think of airplane wings: if the angle between the wing and the oncoming wind is too high, the wing will stall and not give much lift). At high wind speeds, some turbines drive their blades back toward feather; this is option with more predictable behavior. (Other variable pitch turbines drive their blades the other direction to stall them, causing excessive drag, but stall behavior is much harder to predict.) Unfortunately, I feel like I've just confused everyone 
I was mostly talking about the latter case - feathering to reduce power collection and speed in high winds to protect the mill.
But my main argument was for optimizing the blade shape somewhere BELOW the median windspeed, in order to maximize the amount of power you can get from the wind, running the genny maxed-out over a wider range of windspeeds than if the blades were optimized for a higher wind.
The point is that (if you're not running a wind farm where you want to capture ALL the energy you can that's passing over your land) you don't really CARE if you're inefficient when the wind is high. Then you have energy to "waste" (i.e. leave in the environment). But when it's low you need to be efficient, to grab all of the limited resource you can get. So you optimize for a lower speed, and let the mill be non-optimum (but still maxing the genny) when the speed is higher.
You can always throw away extra power (by feathering the blade to make it less efficient). You can't grab more than your blade can grab.
Maybe the drill should be:
- Size the blades to max the genny for most typical wind speeds.
- Optimize the blades for the minimum speed where you can max the genny.
- Make the tower extra strong (because the blades will be a bit draggy in high winds.)
Sound reasonable?
