It's like the sooner one stops smoking, the less chance he has for a "catastrophic" lung cancer; it's not like when you smoke 10K cigarettes you sure won't get cancer but if you smoke 20K you sure do.
Of course, different people call "catastrophe" different things, but the analogy is quite apt here. There is no "catastrophe" coming, in the sense of a sudden disaster. We can stop any time, and the sooner we stop, the better overall result (and less costly) will be.
You are assuming that the climate system can be modeled via linear mechanisms, that human caused changes are more significant (from a climate modeling standpoint) than natural variations, and that the costs of modifying human behavior (e.g. more expensive energy via renewables vs coal) are less than the costs of adjusting to a changing climate.
My biggest frustration with the CAGW adherents is that they don't assign any cost to their proposed mediation policies. Often their cures are worse than the disease.
CAGW is used as a justification for all sorts of increases in government power, which by itself is probably more dangerous than even the worst scenarios promulgated by CAGW supporters.