I guess part of my feeling is, sure, science has problems (every scientist will tell you that) but are there better institutions to give us actionable information in the middle of a pandemic? My feeling is no. So, even if I accept your diagnosis of science as "groupthink, politics and power games", I'd say that's just humans. All institutions, all fields are like that. Nonetheless, science has a pretty good record for delivering more or less solid consensus assessments, so even if it does need some reform, I think it's a pretty reasonable bet that the consensus is more or less on the money.
On the topic of consensus specifically, I don't believe that's actually true. Science is full of people claiming there's a consensus but often when investigated it turns out not to be true, and instead there is an ongoing attempt to cover up or suppress disagreement. A claim of scientific consensus always needs to be checked, unfortunately, although that check can often be quite an easy one of the form "does someone make some well functioning technology requiring this belief to be correct".
There is no requirement that information come from institutions, let alone that it be actionable. If you look at where the most accurate information has been coming from throughout it's been bloggers, skeptics, bands of anonymous researchers like DRASTIC and so on. It's not good and it's not better than having trusted institutions, but all such institutions have been hopelessly destroyed from within, largely thanks to ideology. I see no reason to continue genuflecting at their feet of their smoking hulks. Perhaps better institutions will arise, or perhaps people will learn how to filter information and make decisions without being told what to do. Or perhaps both.
We're at a point where scientific work is the sum of an enormous number of specialists, most of whom are only marginally able to judge the work of those outside their field.
That implies that, to engage with a scientific fact (say, global warming) you are always engaging with a gigantic aggregate of facts from an enormous variety of disciplines.
It's just not realistic for a layperson to absorb all the literature that leads to a consensus and come to an informed conclusion. Sometimes, people who are experts in their fields can do something on that level, but ultimately, they're making those summary judgements at the mercy of implicit or explicit consensus, hierarchy, reputation - all the elements of group-think that make up institutions.
I think what you're describing in regards to consensus is just what a consensus is. Take evolution for example: you can't imagine a more starker difference in views between like, Dawkins or Lyn Marguiles over what that word means. So there's a 'consensus', in that both of them are evolutionary biologists, but a dissensus insofar as what they mean by that. People, in a vacuum, have an infinite variety of opinions, many of which are superficially similar, but are different on a ton of subtle points.
There are a lot of problems in this kind of setup, but I feel if you're going to call institutions in western society 'smoking hulks' when it comes to trust or credibility, mainstream science is one of the last ones you should come to.
Science is often said to be self correcting. I no longer believe this. Rather it's ridden with groupthink, politics and power games. I watched in 2020 as the scientific "consensus" about COVID models became defined by whichever factions were most aggressive, most shameless and most willing to lie, despite the evidence right in front of my face. Like I said, I never developed a strong opinion on the "HIV not the cause of AIDS" theory, partly because I don't really care. COVID is far more important. But could scientists end up collectively believing something that just isn't true at all for decades? Yeah, absolutely. I totally believe that's not only possible but quite likely. Scientists will still be claiming that COVID models worked well when we're all in care homes.
As to why anyone should care: because how do you know who to listen to, if you don't evaluate their past reliability? You say that you'll believe anything academics tell you because it's "not your field". Sounds like you might be a scientist? I think a lot of us would like to know what's true even in fields that aren't our own, especially because governments have a habit of making the pronouncements of scientists everyone's problem.