I remember reading about how Steve wanted to use a weighted average to calculate the accessibility of minerals after a team finds a deposit, but chose not to, because that would mean if a huge but 0.1 accessibility deposit was found, it would ruin whatever deposits already existed there. For example lets say Mars has 20,000 duranium at 1 accessibility. Your survey team finds a massive 100,000,000 duranium deposit at only 0.1 accessibility. Given a weighted average, the huge deposit can sometimes ruin the small one.
My idea is to track the deposits themselves, which would be realistic. Finding 100m duranium in a garbage deposit shouldn't affect the accessibility of the original duranium. For instance, if gold miners in the Yukon strike it rich, that doesn't affect the amount of, or how accessible the gold in those mines in South Africa.
This is probably not a big deal, as I don't think people are usually running out of minerals. Most shortages are due to production outpacing mining, not running out of mining sources. But it would be realistic.
