Immediately decoupled from dividends or traditional metrics like P/E doesn't mean decoupled from the likelihood to achieve good performance on such metrics in the future.
Consider e.g. Tesla in 2013. It had a market cap (directly proportional to the stock price, assuming no secondary offerings) in the area of $2 billion. No one would expect it to pay dividends for at the very least the next decade.
Most people assumed it would go bankrupt, like all other car startups. Some few believed it would be successful. When the Model S launch proved to be successful during the following year, the market cap (stock price) increased considerably.
This was consistent with an increased expectation that the company would survive to the point where it would be able to return money to investors, through dividends, stock buybacks or other measures. But nothing about the Model S launch had any direct effect on dividends or earnings. Looking at the situation today, the company has much better revenues and margins. They could easily choose to have good P/E metrics if their market cap was still $2 billion, meaning that investors who bought in at that price were successful in their bet.
Controversy around stock prices are always caused by disagreements about the expected value of the company's future earnings. Unless it's an exceptional case like Gamestop, where the stock price increases for reasons that are mostly technical. (In this case, the potential future requirement for short sellers to buy shares at any cost).