The answer is a mix of technological expectation, and routine. The industry has steeped itself in twinning "better games" with "better tech", and the biggest gaming publishers aren't some of the most lucrative companies on the planet because they take risks. The end product is forever bigger, bulkier and somewhat bloat-ier, but often within a VERY specific wheelhouse.
For example, Assassin's Creed Odyssey might have added a ton of melee finishers and stealth kills, but its overall supernatural focus is beat-for-beat what Monolith already achieved in Shadow of Mordor. There's no real risk when you're aping one of the most well-received games of the decade. Call of Duty is much the same every year too, and I'll give you a cookie if you can name one unique gameplay mechanic from Horizon Zero Dawn or Spider-Man.
Fundamentally, when the budget is so big - the vast majority being spent on refining visuals, play-testing and frame rates - gameplay NEEDS to make a sizeable return, and minimising anything that could be considered a risk is paramount. Though they didn't have a dog in the visuals race, this "dependable wheelhouse" mentality is exactly what killed Telltale Games.
Read into what happened and you'll find tales of a CEO refusing to do anything other than pump out identikit skin-swaps of 2012's The Walking Dead. Every game after that landmark success felt exactly the same, because the rigid "If it ain't broke, don't fix it!" mentality was applied to every subsequent release.
The problem in Telltale's case feels almost entirely like mismanagement; of not letting a creative team take the necessary risks that would've breathed new life into their games, and them as a studio. The employees have even said as much.
And there's a larger problem with all this "focus on visuals, stick to what works", tunnel vision-thinking: