I've been thinking about something like this.
A lot of people, for one reason or another, want to cut down on their fossil fuel consumption. Let's reduce the question to fossil fuels because that gives you a single metric: gallons of crude used to bring this product to you. Let's also put to one side those who say that there's no good reason to cut down on fossil fuels; at least some of the developed world's population does want to use less oil.
But given a vast array of products, some of which advertise being "green" in one way or another, you don't actually have a good way to know what's the least oil-consuming shopping cart to bring home from the grocery store. Without knowledge, you can't "shop green," any more than you can save money without seeing any price tags. Using labels, news, and PR to inform you how to "shop green" may well be counterproductive. In an ideal world, you could actually get data about the energy expenditure of every consumer product, from bananas to boomboxes. Of course, to do that somebody would need to enforce disclosure requirements on manufacturers and that's not going to happen (and maybe shouldn't happen.)
At the very least, though, it would be a good project someday to make an incomplete estimate, based on what energy consumption data is available from manufacturers. And you might be able to get industry averages or back-of-the-envelope estimates for product categories ("Cabbage is better than lettuce by X gallons of oil per pound")
But this is what frustrates me sometimes. I see people and institutions trying to change their buying habits altruistically (to use less oil or less sweatshop labor or something else) but they have no way to measure numerically how they're doing at their goals, and they often don't seem to care one bit. Why?!?! (Ok, signaling, but grrrrrr.)
Yeah, this is something I struggle with. My grandmother once made a huge effort to find these aluminum water bottles that were theoretically better for the environment and were made locally. Later on they turned out to be made in China, like everything else.
I've also been reading some stuff suggesting that driving a few extra miles to go to a remote but "local" farm ends up costing more oil, since the food produced overseas is shipped in huge bulk that works out to be more efficient.
Oil consumption isn't the only thing that matters, but does make for a metric that should at least be possible to come up with.
Lately I've been thinking about all of the various services and products I consume and how pretty much all of them are bad for the world in one way or another, large or small. Some of the problems associated with them I am less concerned about. Some of them could be construed as good things (i.e. sweat shop labor DOES provide jobs, whatever impact it might or might not have on the overall quality of life).
In general I'd like to live my life having as minimal a negative impact on the world as possible. But "negative impact" is a hugely broad topic and there are a million variables to consider and I just don't have time.
The best solution, I think, would be to have a wikipedia-like website where individual people with knowledge of specific problems can start tagging specific products with the types of negative consequences associated with them, and (somehow) sort those consequences into categories that individuals can decide how much to worry about. Over time it could eventually become a fairly efficient way to track the utility value of things.
I'm sort of hoping something like this already exists, even if in an infant form, and that someone here knows about it. But I doubt it, so the I guess this falls mostly under the post category of "hey someone other than me should devote a bunch of time and energy to this project that I myself am not qualified to do." But maybe a few people here at least have a better idea than I do of the scope of the requirements for it, so the idea can be refined a bit.