I was under the impression that the hydrogen storage (via fuel cell, or compressed air) concept wasn't very efficient?

Correct - but it's a storage method. I wish that I could find it again - but there was a prof in the UK who had a nice setup - wind and PV generation as well as hydro-electric. Now when he had spare energy he charged the batteries (hours of energy storage) and when they were full he pumped water up into the pond (a day or two of storage) and if that was full then hydrogen was created and stored (a week or so of energy).

Now hydrogen does tend to leak out of containers, and it is inefficient creating, storing and then converting it back into energy - but if you've got energy to waste then it's a viable option.

What's nice about being grid intertied is that someone out there will be able to use the energy.

It would be nice if a good energy storage scheme (ultra-capacitors, flywheels, batteries, whatever) came online to smooth out energy from turbines and PV.

But then I hold strongly that we need to pitch the toys and dump the crap. Freezers don't need energy 24x7 and fridges can mostly survive without it. We could also do without fridges. Hot water heaters can easily get by with electricity during only part of the day (but then do we really NEED hot water?) - but energy for cooking is the rub. It's an issue with our food storage. We store a fair bit of food - but we have far more food than fuel to cook it with. We've tried solar cookers and white gas stoves are thirsty beasts and a lot more expensive than an electric range.

I agree on hydrogen storage; it has a quite low round-trip efficiency. Demandshifting and long distance transmission are the low hanging fruit. When it comes to storage, there are many other technologies for storing electricity which outperform hydrogen. Most efficienct are flywheels (Beacon Power has just commissioned a plant to provide frequency regulation to the NY grid using this technolofy), and then utility scale batteries... NaS batteries have been used for years in Japan and are now being installed experimentally in the US, while Vanadium Redox batteries are also making inroads and scale up very well.

I've written more extensively about how to integrate renewables into the grid here

As to Large scale wind farms vs distributed, larege scale is currently favored by the production tax credit. The economies of scale are also an advantage, but this may be trumped in the near future by lack of adequate transmission... small scale is more expensive, but it we can do a lot more distributed generation as we build out the transmission we need.

Small scale installations have the advantage of being cheaper and higher volume. These two properties make a product more likely to be caught by the river of mass production, mass marketing, and mass competition.

I don't know why windmills costs so much damn money, when I've seen dozens of designs that were built from scrap, old hard drives, etc, that were dirt cheap. If someone comes up with a way to mass produce a 1KW wind turbine and energy storage system that pays for itself in under 3 years, it will sell like hotcakes.

Hydrogen production/storage/use may be economically competative using the usual electricity generating tech at ~20-30% efficiency provided the wind power is cheap enough (class 5 resources?) since it could be offered as competition to peaker plants that are at ~15-30+c/kWh based on what I've read. Granted, for the time being programs to encourage efficiency increases in order to not use that electricity in the first place are far cheaper, so that's probably where we'll start, but it's interesting IMO looking at the huge difference between peaker and baseload rates.