Examples?
Here are a few that come to mind right away:
Allowing the coinbase transaction to have the same hash in multiple blocks?
Allowing the miner to create a block that doesn't pay out the full reward?
Requiring the public key to be included along with the signature in the transaction inputs?
I'm sure there are others, but I figure 3 is enough to prove the point.
Allowing unconfirmed transactions to be malleable (one can say don't rely on tx id however the tx id could have been constructed to be immutable from day 0).
Using the use of uncompressed public keys (no advantage over compressed keys and given how critical space in the blockchain is it bloats the resource requirements).
Using 160 bit for PubKeyHash while PubKey only has 128 bit security (to benefit from 160 bit PubKeyHash would require 320 bit ECDA).
Having early coinbase tx payout to pubkeyhash instead of address (just adds needless complexity and now is a legacy cost).
All these vulnerabilities
https://en.bitcoin.it/wiki/Weaknesses#Security_Vulnerabilities_and_bugs
None of this should be an attack against Satoshi, he was brilliant, did what many considered impossible, and solve/prevented entire categories of exploits but he wasn't omnipotent.
I suppose the real Satoshi is more likely to contribute to BTC again than come up with a "great new coin".