Another point further complicating the analysis is that companies which start to do really well simply turn around and reduce the amount of subsequent grants.

Another point further complicating the analysis is that companies which start to do really well simply turn around and reduce the amount of subsequent grants. There is less risk of failure; the expected value of the grant has increased, so the company offers fewer shares to existing employees in the next grant cycle. This effect can be significant as growth companies tend to hire a lot of people (including hot shot VPs with huge price tags) and take more rounds of funding, both of which can dilute initial grants substantially before the liquidity event. The author assumes no subsequent grants, but realistically everybody factors subsequent grants into their back of the envelope calculations.