I'm not sure I follow. My current understanding of what I've read is we would be paying for the electricity used twice. The first time through our electric bill, the second by donating GRC to an renewable energy coin? Or are we trading a portion of GRC for the other coin, so we are holding two coins instead of one in the hopes that one may do better than another?

We would only be paying for the electricity once, with fiat. In the way I proposed the exchange, crunchers would trade their GRC for green energy based coins to offset their carbon emissions; these green energy based coins would also have value, so this wouldn't be a donation. This more closely reflects the second option you mentioned, but the goal is not to hold two in the hopes that one may do better than the other, but rather construct a multi-currency economy where the value of each is increased by the increased flexibility of their utilities.

The current research rewards don't even come close to covering the electricity used.

That's completely true, but we can't act like that's the case, otherwise why bother developing the coin at all? The current value of the coin in fiat is almost entirely tied to the BTC/fiat exchange rate, which is one problem already. Other problems are systemic issues in how our coin is structured, rewarded, and the initial distribution of the coin, which we'll have to address in the future if we want our coin to have value.

If we want to be environmentally responsible a bonus should be given to more power efficient hardware.

More efficient hardware is already given more GRC by virtue of the fact that it is better at crunching. I've written about this extensively, and my most recent proposal enhances the aforementioned effect by removing the disparities between rewards across different projects.

Or a environmental fee assessed and donated to a program that plants trees

I was going to write about this possibility in another post, but it's already been mentioned in the comments - I'm a big fan of offsetting emissions via responsible reforestation and/or preservation of existing ecosystems.

or helps low income households replace inefficient appliances (save electricity and help people reduces living expense) for really old and inefficient hardware.

This would be charity, which I support, but it falls outside the scope of the multi-currency economy I was writing about (although cryptocurrencies which do enable such work could make it suitable - in fact, more efficient appliances save money over time, so maybe people will be able to monetize that somehow).

After reading a few comments I wonder if the energy being used is considered fungible. Most renewable energy has priority for being purchased first by the utilities.

Yes, I'm also struggling with that as well. I'm thinking about modifications/updates to this proposal to resolve that problem.

The way to be green is to use less energy so dirty energy sources are not activated. When we contrive to specifically buy 'green energy' more isn't created, someone else is just buying dirty energy.

There's more than one way to be green. Of course we should use less energy, but we also need more green energy production as well. Regarding your latter point, I addressed that in my comment to @lemouth - that's true in a static environment, but over time, if the supply of green energy can't meet the demand, more green energy will have to be installed.