I think there will be a range of issues from a few diehards hitting serious issues to people just having low-grade issues which they don't notice because they won't be randomizing blocks, effects similar to the hedonic treadmill will make it hard to compare over time, they'll get initial benefits from the usual placebo/Hawthorne/overjustification effects, and subjective self-rating has many known loopholes where you can think you're getting better even as you're actually getting worse - but regardless of the exact distribution or what the worst-cases look like, we won't know for the reasons I list above.
Instead, we'll get another internet circle-jerk about how Soylent is awesome and the critics are wrong.
What would you like to see done differently? You mentioned the more thorough self-experimentation he could have done (really should have done), but there's still someone else who could step up to the plate and do some self-testing.
Thorough studies? Those might also be done some time in the future, whether or not they're funded by Rob (not sure about this point, there might not be an incentive to do so once it's being sold).
Sure, Rob jumped the gun and hyped it up. But most of the internet is already a giant circle-jerk. Doesn't stop people from generating real information, right?
Rob Rhinehart's food replacement Soylent now has a crowdfunding campaign.
If you're interested in one or more of these benefits, send in some money! There is also a new blog post.