My feeling on the matter is that you don't go there until you have, at minimum, a well developed concept that you're ready to start moving on, and it's probably

My feeling on the matter is that you don't go there until you have, at minimum, a well developed concept that you're ready to start moving on, and it's probably preferable to have gotten started on a MVP. The other question you should ask yourself is whether you need to be located in Silicon Valley at all. As you note, there are both personal barriers to entry (i.e. expense) as well professional barriers to entry (i.e. it may be easier to gain initial traction in another market that isn't as inundated as Silicon Valley is with other startups), and every case is different. Like most other decisions, it really should just come down to a cost-benefit decision. Are the additional costs of moving to Silicon Valley offset by the benefits (better networks, available talent pool, availability of VC funds), or are you better served somewhere else? Only you can make that decision.