Just to give an idea of what the actual trade-offs are between the benefits of wind and the immediate ecological impact, I ran some numbers a few months ago and reached an estimate of somewhere between $1700 and $2500 saved in "social cost of carbon" per dead bird, for on-shore wind power in Scotland and the north of England. Social cost of carbon is the estimated value of the benefits of reduced carbon emissions, including benefits to human health. For a non-endangered bird, even $1700 saved is a good deal. Clearly a pigeon's life is not worth $1700 (not that pigeons are particularly at risk from turbines). For a critically endangered bird, it is not a good deal at all (extinctions are very bad). At some point there's a cross-over.

The problem is that as you build out wind power, you kill more birds, which starts putting populations under stress that weren't stressed before. In other words the cross-over point moves. That makes effective mitigations more important the more wind power gets built. Ideally a doubling of capacity should be accompanied by a halving of the kill rate per unit of capacity. A mitigation strategy that could cope with all the planned capacity increases would probably look like: paint jobs on all new build, plus effective radar that can brake the turbines, plus a form of active deterrent (i.e. microwave or sonic repellent). All this while policing the wind companies by tracking tagged birds.