Unless you think you're a better judge of company valuation than the market, the current risk-adjusted value of the company is approximately the valuation set by the recent round of funding.
The fact of future dilution is irrelevant, since this impacts the investors as much as the employee.
This analysis does ignore liquidation preferences. It also ignores the fact that the options have a non-zero strike price (though probably substantially below the preferred stock with those liquidation preferences). But in the end, startup outcomes tend to be binary: very high or nearly zero. In both of those cases, liquidation preferences and strike prices wash out in the noise.
I agree though: equity is almost the definition of a mixed bag. That's why, as an employee taking substantial compensation in equity, you need to take an investor mindset in choosing where to work.
I don't understand either of your first two sentences.
The 50% risk thing came from you: you suggested calculating based on a 1 in 2 chance of successful outcome. If you're forecasting based on a 1 in 2 chance, you divide by two, right?
Second: as an employee, what do I care whether investors take a haircut? We're computing my outcome, not some notion of fairness.
The point is, if investors are in at $2M, we are assuming $2M is market price. The point of market price is that includes things like "google will compete with us" or "my share will be diluted when the company raises again" as well as "the company might fail".
Now, since you personally are investing in the company (albiet with your labor rather than capital, but it's still an investment since you're paid in equity), you have to decide if it's worth investing in this particular startup. You do that by thinking about whether the startup is going to succeed or fail, since outcomes are pretty much binary.
I disagree, though, that startups have binary outcomes. In fact, one of the big problems in "investing" in startups (cash or labor) is the misalignment of incentives. It's common for the founding team to pursue outcomes that will enrich them but zero out common stock holders, and it's common for investors to push back on exits that would enrich everybody but not satisfy fund goals.
The fact of future dilution is irrelevant, since this impacts the investors as much as the employee.
This analysis does ignore liquidation preferences. It also ignores the fact that the options have a non-zero strike price (though probably substantially below the preferred stock with those liquidation preferences). But in the end, startup outcomes tend to be binary: very high or nearly zero. In both of those cases, liquidation preferences and strike prices wash out in the noise.
I agree though: equity is almost the definition of a mixed bag. That's why, as an employee taking substantial compensation in equity, you need to take an investor mindset in choosing where to work.