Make good decisions. Join the right team.
-- Either be a founder if you want to be there in the early days.
-- Or join a "sure thing". EG: Google, Twitter, Facebook about a couple years before they went public were already household names and really well known.
I don't know how much upside you get joining a sure thing like that, but that's how you make sure your options will come into money.
Being employee number 5-100 of the average Silicon Valley startups is a losing proposition because the risk adjusted value of your options will never compensate you for your lost salary. (especially if you have to live in California- you're better off working for a startup in Austin than California due to the cost of living and tax situation. The higher salaries in California don't cover the difference.)
And yes, blame me, I turned down being employee number 13 at what became a $6B enterprise software company. Would have been CTO or way up in the executive team because they were a bunch of biz guys who needed a hacker. Instead I worked for just a year for a small business (not really a startup this was before "startups")
But it's damn hard to tell the difference at those early stages.
And when questions like "what's the total number of shares outstanding on a fully diluted basis?" (back when companies would say "You'll get 10,000 shares!!!!111!!") are met with "sorry that's confidential" during the hiring process, it is a bit difficult to do proper due diligence.
(I started at Google in 2004 and did pretty well.)
There is another facet to your second point, I think. Even if the upside from the "sure thing" two years before they IPO is not great from a financial perspective, the career growth is much faster than joining a mature mega-corp. For example, many people who joined Google/FB/Twitter a year or two before the IPO are very senior at those companies now. Granted, not everyone who joined Google in 2004 is a VP, but the probability of rapidly growing your career as the company grows is much higher than for someone joining Microsoft in 2004.
That fast career growth leads to either 1) High compensation ten years later as a director/VP or 2) Exec roles at sure-thing unicorns with meaningful equity, should they choose to leave Google/FB/Twitter.
I think there are two justifications to go early stage if you care about the financial aspects:
1) Be the founder, as you said
OR
2) You have concrete reasons to believe that the company has a meaningful competitive advantage in a large market.
These examples are far and few. For example, Google in 1998 had a meaningful competitive advantage, but Facebook in 2004 certainly did not. WhatsApp in 2010 did not, but I'd argue with Carmack, Oculus might have had (although unsure in this case). As you said, it's really hard to say at such an early stage.
Of course, if financial upside is not a major concern, then impact, agility, working on interesting stuff, avoiding mega-corp red tape, are all valid reasons to go early stage.
For instance, when youtube got bought, VCs were all interested in investing in online video companies. At that point, though, Youtube had already been bought! They were like 5 years too late.