[whack]

> Lots of global warming sceptics point to some subset of experts who are themselves sceptics

Yes, this is exactly why you should:

1. Consult more than one expert, until you've found a clear majority. And then trust in that majority

2. Pick either a couple experts who are the most highly regarded in their field. Or use random sampling to pick a couple experts from among a pool of equally qualified experts

As you yourself pointed out, people who fall for fallacies often form their own judgement first, and then cherry-pick experts to lend more support to their argument. This is precisely the danger in telling lay people to trust in their own judgement.

Cherry-picking of experts is the opposite of expert-guided decision making, because the cherry-picking part is being driven by lay judgement, not expert recommendation. A truly expert-driven approach would rely either on the most highly regarded experts, or random sampling from a pool of equally qualified experts. This leaves minimal room for cherry-picking and confirmation bias.