Yes, the answer is simple, and you are correct.
But the cost differential is laughable.
A 5m machine of 25% is cheap to build. A 34% 3.2 will all the gadgetry is hyper expensive. Two towers, two controllers, two dump systems, 2 sets of bought blades ( you claim home built blades are inferior ) 2 clippers and the list goes on.... and more real estate to put them on.
$500 gives me a 5m machine.... 22-27% depending on how I match the blades to the expected wind regime .......your up for $1600 for mppt, before you even start to build your two machines with their two towers, and two sets of blades, and two sets of clippers..... and they will probably fail within 8 years anyway.
I haven't seen inside these things , but I suspect you will find a reasonable sized cap bank suffering ripple. They will fail by design... just like batteries will. The KOH will eventually cause failure as in ALL modern electronic stuff with PWM power stages instead of the old reliable transformer.... from your computer , your dvd recorder, satellite transciever.. you name it. It cannot be gotten around. I think the best that the worlds very best manufacturer can possibly guarantee is in the 8 year range. Electrolytic caps are most fallible, and you will do well to keep them as cool as possible to prolong their useful life. They will last 16 times as long at 60C than at 100C.
The electrolyte is the main cause for the lifetime being limited for the cap. It must continuously modify it's electrical characteristics at the dielectric layer, and a combination of self healing and drift through the seals will sap the electrolyte down. Heat will make the electrolyte less viscous, which will actually help lower the ESR. The current flow is mediated by the ionic nature of the KOH, and as the ESR lowers, higher current may be seem in the caps.
The ripple forced on the caps by the buck switching (superimposing the switched DC on the background DC) is the culprit. It cant be gotten around, any more than the evil death of batteries can't be gotten around. Both systems can be designed to be less abusive to the batts and caps, but only by never discharging the batts too far for too long, or keeping the ripple to a very low level so that the caps electrolyte gets an easier ride.
The estimated lifetime can be empirically guessed at by the use of temp time graphs from the various manufacturers. Temperature and voltage will help accelerate the degradation of the electrochemical characteristics of the capacitor .
So you have built in death in your 2x3.2, and expense I can't comprehend. Living where you are in contact with the current manufacturers helps, but where i am... forget it. It would cost $500 just to send it back.
So I won't be swayed by bells and whistles either. I still applaud you for doing it, but I would still caution anyone else particularly if they are not in the USA to either build their own, or stick to tried and true methods.
I'm not sure the comparison was useful.
...............oztules
Edit: It is interesting to contemplate what will happen on about 5-8 years hence with all the grid tie inverters that are out there. They have been installed in this country extensively over the last 2 years or so.... there must come a time where they are all going to fail together. The cap failure rate is a bathtub graph... a sample will die within days, then drop to a long period of only scant random failure, then toward the expected life end, it will rise up again very steeply then taper off. There will be piles and piles of grid tie units on the tip.... will be interesting times indeed.
I suppose to ones mounted on the sunny side of the houses will go first.
