The graph shows 50 companies (they don't do a good job at scaling the site, but the data is there).
Netscale and Amazon are only anecdotes and pure exceptions of the dot.com era. The set of companies that went burst during this period because of bad management is composed of at least 15 publicly traded companies. If your compensation included shares of any of these quick IPOs (AOL, Yahoo, pets.com, Global Crossing, etc.) your shares would have ultimately be zero as well (compounding interest is what makes you wealthy with these long-term horizon packages and you wouldn't have sold all your shares at ipo).
Yes those 50 companies are very recent in SV history. Hence we need to look back farther to see the change, because our standards were set many decades ago.
And whether a company goes bust post IPO isn't really the issue here since liquidity is what we're discussing.
Netscale and Amazon are only anecdotes and pure exceptions of the dot.com era. The set of companies that went burst during this period because of bad management is composed of at least 15 publicly traded companies. If your compensation included shares of any of these quick IPOs (AOL, Yahoo, pets.com, Global Crossing, etc.) your shares would have ultimately be zero as well (compounding interest is what makes you wealthy with these long-term horizon packages and you wouldn't have sold all your shares at ipo).