These examples don't actually disprove anything about microeconomics though.
Your example of the fish markets is just a good example of the fact that people have varying marginal utilities of money. Of course people try to get the highest price they can get - taking into account, of course, the time they'll be receiving the money. Obviously people will see if they can raise the price, but they're still in competition with other sellers; the price cannot be raised indefinitely. And it's not always clear whether other factors have changed - things like fish supply and catching are notoriously dependent on the weather, demand for fish varies seasonally and with changing fashions; it's a perishable good with not-super-fungible markets - maybe the prices know something about fish that you don't.
You give the example of super-high pricing for drugs such as the Daraprim case with Martin Shkreli. Note that this market is overseen and hugely, hugely regulated by the FDA, which imposes very lengthy and costly approval processes on drugs (even different manufacturers of off-patent "generics".) More generally, these drugs are on-patent - both of these situations are a de-facto monopoly that grants exactly the kind of economic rents that microeconomics predicts.
Ecommerce pricing also doesn't disprove anything. You could, for example, explain this by other factors, such as that ecommerce warehouses and logistics chains are capacity-limited, so "surge pricing" would make good sense to smooth demand peaks. But that's besides the point, really, because microeconomics would absolutely predict that if people, on average, have a preference for shopping at a particular time, then a seller may attempt to capture this preference in the price. This isn't done in brick-and-mortar retail, since changing the prices hourly would be a huge hassle. But in ecommerce, it's easy, so prices modulate with demand. And stores have always changed their prices over time, it's just that the internet makes it easy to go for a smaller granularity.