If you leave within 2 years you never have to experience the long term effects of your decisions so there are a lot of valuable things you never learn.

That's not true of the Facebook people I know. They include some of the best engineers I've ever met. (No other connection to Facebook, just a personal judgment based on their externally visible work.)

More generally I find that different people work effectively at different timescales. Some of the big movers in the industry have achieved phenomenal things by focusing on one problem for decades. Definitely true for databases, which is what I work on.

Boring isn't a bad quality for an engineering organization. Chasing trends and doing "cool" things and trying to be clever tends to not create good long term engineering products.

Wherever I worked, mostly the more competent people were the ones leaving. Isn't that to be expected?

Some people go to work to get paid, and are less interested in dazzling others. (Facebook strikes me as a great option for such people, by the way.)

It would be fine with me if someone considered me dull; I decided long ago that I care less about being interesting than being interested.