Also, some of the frighteningly hard to stop feedback mechanisms have started, such as glacier melting has hit some extremes that scientists feared would be points of no return. Whether this will play out in the ways of those 1970s worst case predictions we don't know, and a variable here is that we are working to avoid them (badly, but still some avoidance work is better than doing nothing).
Some of that is like the Y2K problem: the self-unfulfilling prophecy. We've warned ourselves for decades what the risks are and what the worst cases are. We've put in a lot of work to mitigate/slow/stop the worst cases. We've partly failed, several times over, and had to invent new worst cases. We're still working to prevent those new worst cases. We likely will not ever hit a worst case, we're just in our own feedback loops trying to find the acceptable compromise. If the acceptable compromise "turns out alright" people will claim we never really were in danger of the worst cases, discounting the huge amounts of effort we put into avoiding it. "The prophecy was self-unfulfilled so it wasn't really a big deal was it?" "Y2K didn't happen, so why was everyone so worried about it?"
The really, really bad scenarios depend either on the CO2 emissions not reaching an inflection point and decreasing (which I think will happen, but how quickly it happens will have a big impact on outcomes), or on some positive feedback mechanism kicking in, of which there are a few potential candidates but a lot of uncertainty for where the thresholds are (predictions range from 'already passed' to 'in the relatively far future'). These should neither be ignored as a possibility nor treated as inevitable.