Thanks for the followup, much more well formed. To a few specific points:
> the slope of the employee gain once the preferenced shares have met their obligations will be substantially higher than in my math above.
Yes. This is all but required by "the rules".
> subsequent rounds increase the number of outstanding preference shares
Yes, but by lesser and lesser amounts of dilution, and at higher and higher valuations. Ignoring down rounds, which are much more common than up rounds. But here I think we are talking about a "successful" exit, which as rare as they are, in the majority still leave option holders high and dry anyway.
> your employee strike price could have been set at a time when all preferences would be satisfied, The math will then shift back to a 6x required gain (for a strike price equivalent).
True, however in these cases almost everyone (actually everyone, AFAIK) receives RSUs, not options. RSUs have an actual, current value, and do not require any appreciation in price at all. In fact, a loss in price is still valuable to the holder. You can't go underwater with RSUs. (And you can't face uncertain taxes; be forced to make awful decisions; etc.)
> On a 10-year liquidity horizon,
This is the real problem with equity compensation today. We've outgrown the system under which options were a great incentive. Golden handcuffs, that most exceedingly often, turn into fool's gold. Honestly, it's time for YC to insist on 10 year termination-exercise rights. The appeal of ISOs is still there for early stage companies, but AIUI one can convert from ISOs to NQ's after the 90 day legal requirement, without penalty. win-win.
With a 10 year rights, companies are forced to offer sufficient refreshers to keep staff. This is completely fair and will push companies to earlier exits. On the employee side, people need to insist on it as well ...
I don't see it happening because it destroys the VC business model.