Tis true. It's a conundrum. Do you give power to the regulator which will help them stop the bad guys, but also enable them to be bad guys at the cost of the good companies, or not?
Free market folks argue that the market can figure out every one of those things you mention. But most of them understand it isn't a free lunch.
It is a conundrum, and it is one of the hardest political problems for humans to solve. It's half the reason we have politics.
I generally think that a case-by-case approach, with clear avenues for both rulemaking and redress, with democratic (direct or representative) control is a good starting point for solving this conundrum.
I allude to the problem of finding market solutions for ethics problems in the last part of my previous point. Markets do not optimize for ethical behaviour. Largely because of information assymetries, largely because people are not perfectly rational agents, largely because of the power imbalance that exists in many of them, and also largely because people are poor at measuring specific examples of long-tail risk.
Good rules also reduce market friction, because they let me make reasonable assumptions, like 'This product is probably not going to kill me when used as instructed, because it adheres to X, Y, Z objective standards', instead of having to go down an endless rabbit hole of 'Buyer beware, do your own research!'
I would go a step further and argue that markets optimize for institutions that will seek to consolidate monopoly power and destroy the freedom of the market in which they formed. They specifically optimize for unethical behavior, as a natural consequence of the incentive systems in place. I argue that this is true even if there is not a large bureaucratic government to take advantage of for regulatory capture.
Basically, classical economics is predicated on a knife-edge equilibrium that ironically needs government intervention in order to prevent people from deviating.
This is speculation and pedantry, but: What if they do? What if they already weed out 90% of it, and the remaining stuff that you notice is the optimal amount given the cost to weed it out (no matter how said weeding is done).
Pedantic counter-example and Exhibit A: The endless litany of unraveled crypto scams that keep getting bigger and better, year over year. There's one that's going bust right now.
It's the poster child for a perfect market environment, where everyone involved is simply trying to optimize their returns, and there are next to no rules. It's also full of hucksters, thieves, conmen, liars, and flat-out frauds, to the utter detriment of any legitimate commerce in the space.
You don't think the market could sort this out through time? Free market believers don't think the market will sort out every problem overnight - it is pretty much the belief that the bad players get shaken out over time.
The harder ones seem to be healthcare (massive info asymmetry) and anything related to pollution (externalities). The market has had time on those and not done so well.
Even if they did, it would be idealized free markets that did that.
Not the very much non-ideal, largely non-free markets we have, rife as they are with information asymmetry, non-commodity products, buyers with urgent time needs, and a host of other aspects that utterly invalidate the Free Market Will Solve Everything gospel that so many try to preach.
Free market folks argue that the market can figure out every one of those things you mention. But most of them understand it isn't a free lunch.