It's a sucker's deal because point-zero-whatever percent of some nondescript startup is not likely to compensate for the lower wages and benefits relative to larger companies. And there certainly are jobs where you can trade lower pay for higher stability (i.e. govt). But no one in the US tech industry (big or small) ever claimed to be about lifetime employment in recent times (well maybe some crazy outlier firm somewhere does). Your long-term financial stability is your responsibility, not theirs. The tech giants will have their layoffs sooner or later too.
And if you're going to cite FAANG salaries and recent stock performance, then to be fair you need to compare it (also utilizing hindsight) to the fastest-growing startups with the best stock performance.
I'm not sure this is the case, since if you made an employment decision, say, 1 year ago (or whatever), FAANG was a known quantity back then, too.
You wouldn't know this year's FAANG performance, but you could arrive at reasonable-ish expectations based on previous years' performance, whereas a startup that was just getting off the ground at the time would be a total unknown even if it managed to make everybody rich in this past year.
As is common knowledge, though, the startups that make everyone rich (let's ignore the moderately successful ones for the moment because we're talking about taking reduced pay in the hope of future stock performance) are the unicorns, so statistically you should not be expecting that yours will (particularly if you're an employee who doesn't have the business-level decision-making power to significantly impact that probability).
Looking at some price curve and trusting it because of a trend in hindsight is just bad logic and dangerous investing. Those curves change quickly and unexpectedly, crashing right at the time when the most people become bullish believers due to being wrong so often otherwise. You can only trust fundamental reasons for investing, and those are way out of whack currently. I see the same story again every ten years and we are overdue.
My point isn't that the curves will stay steady or increase, but that you have historical data on them to draw an informed conclusion, bullish or bearish, which just isn't the case with startups.
In terms of risk, they are completely different animals. FAANG are included in the S&P 500 index, which automatically triggers investments from many funds that have rules about this sort of thing. Many high-growth startups do not have that going for them. Further, they often fail to sustain their meteoric growth for sustained, long periods, adding further imponderables to the mix. Having been at many such companies myself, and having friends in FAANG for roughly the same period, I find that on the whole, they are doing much better money-wise, and surprisingly, in terms of career development.
In terms of salary compensation, again, growing small-mid size companies again fall short in comparison to the FAANGs. The reason is that FAANGs have good structures set for performance management, career paths, and compensation increases. The former, in contrast, are almost always lacking in having good structure in these areas. So for the average engineer who's mulling a choice between the two, the risk/reward ratio tilts towards the FAANGs.
Note that I'm mostly talking about money, and to a much smaller extent, about average career growth here. There are still plenty of other good reasons to work at small-mid size companies, but in terms of money, it is difficult to argue that going to the FAANGs leaves you worse off than the alternative.
Most count their FAANG stock with 0 growth. You still get 180k for junior, 250k for mid level and 350k for senior per year.
It's never a good idea to take a pay cut (salary-wise) to work at a startup. Moneywise, it makes even less sense to leave public company stock (esp. FAANG whose stock is growing at a healthy clip with much lower risk) and salary compensation to go work at an early stage startup (e.g., pre series-D, let's say).