I'd argue that where RethinkDB fell down is on a step you don't list, "Understand the context of the problem", which you'd ideally do before figuring out how many people it's a problem for. Their initial idea was a MySQL storage engine for SSDs - the environmental change was that SSD prices were falling rapidly, SSDs have wildly different performance characteristics from disk, and so they figured there was an opportunity to catch the next wave. Only problem is that the biggest corporate buyers of SSDs are gigantic tech companies (eg. Google, Amazon) with large amounts of proprietary software, and so a generic MySQL storage engine isn't going to be useful to them anyway.
Unfortunately they'd already taken funding, built a team, and written a lot of code by the time they found that out, and there's only so far you can pivot when you have an ecosystem like that.