Everywhere you look somebody is hyping green as if this were a key computing requirement. It isn't - and my advice is simple: don't bet your career, or your company, on this movement having legs.
Instead, be absolutely sure that any action you take is mainly about saving money - even if you have to sell it to others as about saving the planet.
Basically: it's great to save money on power costs, but if you make changes, make them for the long green, not the fad.
Look especially carefully at the sensitivity your net present value calculations on green driven facility, hardware, or software change has to guestimates for future power costs. What happens to your numbers if McCain's energy policy gets implemented in the United States and the huge cash flows currently supporting development in places like Brazil and Kazakistan suddenly reverse direction to flow into places like Ohio, Utah, and Florida instead?
If your numbers show any significant sensititivity to downward change in future energy prices then the bottom line is simple: shelve the project until it's clear which way the market's really going to go.