Facebook ran an experiment where they intentionally crashed their Android app to discover the threshold at which users would give up and go away. But someone familiar with the experiment said: “The company wasn’t able to reach the threshold. People never stopped coming back”.
https://www.theguardian.com/technology/2016/jan/05/facebook-...
I think this is a classic case of the Innovator's Dilemma: you start by making something worse than existing products on the market but that completely address the features your customers care about most. Eventually your customers demand more features and you have to turn a profit, so the product gets bloated and the p90 experience suffers.
Which almost feels ironic given that they tend to hire some of the best designers around the world.