Actually, I work for a small tech company, and have to work around other people's undocumented interfaces all the time just like everyone else in this industry.
Your vision sorta works for software-based startups which are not capital intensive and dominated by labor costs. And I've seen first hand how telecommuting overcomes bad immigration law, allowing people in the west to cut the deals that cannot be done remotely on effectively behalf of foreign workers. Bad immigration law is indeed a government-induced problem, and this is one of the best workarounds we have.
Even then, the idea that no good local education is needed and that some body can just get a laptop, read some docs, and start making money is too optimistic.
But the larger problem is I don't think there is enough demand for YC-type stuff to significantly change the standard of living of a state. Consider how much money is dumped into SV startup before it makes money on average, and how few people SV employs due to the productivity of tech. This is not a good recipe for an entire nation.
The classic export goods are material goods; the far simpler unit economics ensure a steady stream of money back home, and the produced goods can (at least in principle) also be consumed domestically to make increasing in the standard of living not import-dependent. No country has gotten rich in the modern era without some of this, and I don't see this trend changing.