Why do people keep obsessing over this "reliability"? What matters is whether the turbines are net energy positive. (Unless of course you're just obsessed with the idea that cheapest is always the best, which is fine for now but is kind of crossing your fingers for the future.)
Yes, you need to pair them with gas plants.
Yes, the paired gas plants are less efficient than a CHP can be.
But as long as energy captured more than offsets the efficiency loss it's a winner.
Also, studies have indicated that as you add distributed renewable capacity you decrease variability. If so, then, on average, fewer of the inefficient peakers will be running, meaning that increasing renewable capacity makes renewable energy more efficient.
These variable sources aren't the full solution to future energy problems, but if they're able to increase usable energy then they're buying more "easy" time now and, unless we actually come up with we'll need them as part of the energy generation in future then in the future will provide some of the energy that will be stored and used.
The big problem with wind is not the reliability: it's building the transmission capacity so that you can distribute it effectively. That's also expensive, but tough shit. Unless somebody can tell me how you're going to keep electricity supplied for the next 1 billion years using _present_ technology (no, thorium isn't ready yet) we need wind, solar, tidal, hydro, digesters, biomass and every other capacity-limited or unreliable energy source to be developed so we can continue to do all the fun stuff until we get killed by a passing galaxy.
Complain about the implementation, priorities and corruption, but don't complain about the limitations of the technology, because that's not the real problem.