Let's be honest, TCP/IP shouldn't be what we use. Since the government held a monopoly on the early Internet for the first half of the Internet's existence they got to embed whatever they thought worked best instead of what actually worked best.
Believe it or not, that probably helped development. If people discover a tech that's logically flawless and a stroke of pure genius, they'll get awestruck and fall into an echo-chamber where the development is confined to hermeneutics. Kinda like the clonecoin scene...
Having a starting tech that's kinda clunky means crowdsourcing kicks in. People see the clunk, think of improvements, get their hand dirty and ignite a chain of innovations that grow the tech into something that none of the pioneers even thought possible.
Fact is, disruptive technologies that make an impact on the world are usually nursed by bleedin' amateurs or pros from a different county (so to speak.) The old "enginer" of the early railways days was somewhere between a skilled mechaninc and a sure-hand do-it-yourselfer. Real engineering came in later. Kinda like the PC...
Now, I'm not trying to deny Satoshi Nakamoto's genius: we all know that the blockchain is historic. But the fact that his original code was kind-of obscure meant that the early devs had something they could really get their hands into wrt improving it. Its very obscurities proved to be the kernel that got Bitcoin on the innovation train.