Market-for-lemons is sometimes a useful framing, but there's such a thing as taking it too far.
Markets, even markets for lemons, are frequently in non-equilibrium states. That can happen because actors have imperfect information, because underlying conditions have changed, because some actors are new to the market and haven't settled to equilibrium yet, because people have personal beliefs that override market-rational behavior, because market actors are simply bad at their jobs, or any number of other reasons. Far more information-perfect, rational, and optimized markets than hiring frequently show inefficient behavior in practice.
Yes, there are market pressures that tend to remove good candidates and good jobs from the market, and that incentivize both sides to pretend they're better than they are. But a tendency is only a tendency, and not a guarantee, and incentives can be missed or ignored or defied or failed in implementation. A bad company might be incentivized to lie to you, but they have to be self-aware that they suck (a trait terrible companies very frequently fail at) and capable of implementing the deception, among other things. And that means that your impressions when you speak to them do carry meaningful signal that should shift your beliefs about them.
A new job might be a lemon. But it also might not. You don't need to be certain that it isn't; you just need to be sure enough that the expected value of taking the new job, minus whatever risk discounting you're choosing to use, (minus the cost of searching for one), is greater than the expected value of staying in your current job.