I agree with everything you say, except for the part about discussing global collapse limits being deck-chair-ish. We are on a path to a lot of warming, where many of these local tipping points have and will become a reality. Having an estimate on when these effects combine to form a global tipping point is essential. It is essential because it informs how hard our policy response is. It is all well and good to try and aim for <2C warming but if we are obviously overshooting that mark, we need to know how far we can overshoot without passing a point of no return. If the consensus became that the trigger point for global collapse was imminent(say less than 10 years), we could plan a response that guarantees emissions reductions in 10 years regardless of the suffering caused (e.g. austerity to reduce energy use without consideration to poverty, banning ICE vehicles for non-essential purposes, sanctions for fossil fuel use). If we are incorrect and the trigger actually occurs in 100 years, we have just caused additional suffering by not planning our response longer. If humanity has an imminent existential risk, doing horrible things becomes morally justifiable (e.g. putting 4 billion people back in poverty). If we are 100 years off collapse, putting 4 billion people back in poverty is immoral without considering longer term solutions. Which path you choose is determined by our predictions for the future. Optimistically planning for <2C warming is good but unrealistic based on current trends. Unfortunately cheap energy is the basis of the world economy. Thus all solutions cause suffering for someone, whether by increased food prices, medicine prices, loss of jobs. Of course this has to be compared to the harms from climate change, but how can we make this comparison if we are ignoring the worst case outcomes (which I would argue are the more realistic cases)?
The fact that neither of us can find information on global tipping points in IPCC documents is to me a sign that the political aspect of the solution is not being done well. The IPCC essentially says "keeping warming <1.5C is essential" and "we are expecting more than 1.5C warming based on current trends" but then they make no estimates of the existential limits for the trajectory we are expecting. It mirrors my experience with environmental regulation. Policies where you can pollute, as long as the pollutant is kept at < X% concentration. But why X? What happens at X-1 or X+1? That is how engineering problems are tackled, but political solutions aren't as systematic. X is likely choosen not because that is the best outcome for society, but because a committee arbitrarily agreed that it is achievable. As we've both seen, the current IPCC limits are based on local collapse of systems such as coral reefs. But these systems are already being affected and are not expected to recover (based on the status quo). If that is the case, I think we should shift our focus from estimating the most warming we can have without _any_ negative effects (unrealistic) towards policies based on cost-benefit (pragmatic). Decarbonization in 10 years looks very different to decarbonization over 100 years. I'd prefer 10 years, but I don't want to ignore the people we will harm in doing so and I want to know what happens if we can't get our shit together in 10. We cannot make an informed decision unless we have limits which are not arbitrary and are realistic.
Edit: I thought I'd add a response to your bridge analogy. I think our current situation is like speeding towards a failed bridge where we can either ditch off the road down a cliff _now_, or go around the corner where maybe we can ditch into a bush. We need to know what is around the corner before we decide to go off the cliff. A real-world example for this: biofuels were pushed as an alternative to fossil fuels via subsidies, which caused minimal benefits to global warming but caused the worsening of starvation in the third world. If we had waited until better solutions were available or bad solutions were 100% necessary, we would have avoided suffering through starvation. Reflexively saying "we are in an emergency, quickly try solution <Z>" isn't useful unless we are sure that we are in an emergency (read: existential risk) AND we have confidence in solution Z improving the emergency. Doomerism breeds bad solutions and unnecessary suffering. Doomerism might be the right response, but the IPCC reports do not convince me of that.