If we spot a planet killer on collision course for next year, there's not much we can do about it. But if orbital monitoring or modelling can tell us a rock has a 50% probability of crashing into us 20 years from now,then we've got time to come up with a way to deal with it, and start a program to get it deployed. Fortunately, Newton's Laws are sufficiently well understood that this kind of prediction is well within our capability now, providing we have the data to support it. As it turns out, most of the potential threats to earth will be much more likely to be in the category of "long term future threat".

This is precisely the type of project that should be planned "for eternity". The initial segments might be useful for limited objectives like evacuating cities before they are hit, but in the long term, we will need years or even decades to implement a full planetary defense. The initial investments should not be a one-shot deal, but the first installments of an ambitious and integrated system planned for the long run. We need to spot these guys many years before they hit us, not just a few months ahead of time.