Yup, there was a bug that allowed the attacker to create an unlimited number of Bitcoin, it was fixed, but there seems to be this myth that SN provided complete bug free working code from the outset.
Yes, but implementation errors-- not fundamental flaws in the theory or underlying mathematics.
It's not a statement about such things being impossible, just unlikely to come from a single individual working in complete isolation up until the bitcion whitepaper release.
Additionally this alone would be merely peculiar on it's own, coupled with the lack of retrospective investigations uncovering -anything at all- significant about the person moves it to suspicious. Adding in the subject matter of cryptography + pseudo anonymous money and it strains credulity not to consider
You're placing way more value on this than makes sense: Bitcoin makes use of good primitives, but no better than any expected familiar with the SoTA in the late 2000s would have selected for a greenfield project.
Maybe the most unusual primitive selection in Bitcoin is secp256k1 for ECDSA, instead of one of the more common NIST curves. But even that is understandable, given that Nakamoto was active in the cypherpunk community and concern around the constants used in the NIST curves was a common discussion item at the time.
As far as I know, there's no concrete evidence that the NSA has compromised the security of the NIST curves. That would be weird for them to do, since they use those curves internally to encrypt data classified at Secret and higher.
Bitcoin had to be forked in the early years due to critical errors in the original code. There was nothing perfect about it.