This is a very valid point, and well taken. However, the criteria for a lunar mining colony are different. There is no need for a completely sealed, self-sustaining ecosystem. Assuming we can solve the problem of getting O2 et al out of lunar soil, then there's no problem with tweaking the gas levels as needed. One hopes that there will at least be *some* sort of bio-cycle handling much of the C02/O2 turnover, as well as providing food and helping with waste management... but there's no reason you can't add more O2 at some point as needed.

Now, when you start to talk about permanent and more distant settlement colonies (ie Mars) then you really want to close the cycle further. Besides, regardless of space exploration, we should continue to try to understand ecosystems by constructing artificial ones. What better way to learn about complex interactions (which we're affecting in poorly understood ways) then with simplified models? So Biosphere++ in any case....