I was reading the cryptography mailing list when I came across something very interesting that dates back to November 2008. According to Wikipedia's article about Cypherpunks, the post I am about to share below was the first answer that satoshi got about Bitcoin.

Cypherpunker James A. Donald was the first who noticed what Bitcoin will have problems with scalability. In his reply, he wrote:

To detect and reject a double spending event in a timely manner, one
must have most past transactions of the coins in the transaction, which,
  naively implemented, requires each peer to have most past
transactions, or most past transactions that occurred recently. If
hundreds of millions of people are doing transactions, that is a lot of
bandwidth - each must know all, or a substantial part thereof.

There are lots of proposed solutions, lots of hardfork made to increase size yet many still uses bitcoin (1mb), now we used segwit to reduce much more transaction data (separate signatures in transaction data)

Long before the network gets anywhere near as large as that, it would be safe for users to use Simplified Payment Verification (section to check for double spending, which only requires having the chain of block headers, or about 12KB per day.  Only people trying to create new coins would need to run network nodes.  At first, most users would run network nodes, but as the network grows beyond a certain point, it would be left more and more to specialists with server farms of specialized hardware.  A server farm would only need to have one node on the network and the rest of the LAN connects with that one node.

The bandwidth might not be as prohibitive as you think.  A typical transaction would be about 400 bytes (ECC is nicely compact).  Each transaction has to be broadcast twice, so lets say 1KB per transaction.  Visa processed 37 billion transactions in FY2008, or an average of 100 million transactions per day.  That many transactions would take 100GB of bandwidth, or the size of 12 DVD or 2 HD quality movies, or about $18 worth of bandwidth at current prices.

If the network were to get that big, it would take several years, and by then, sending 2 HD movies over the Internet would probably not seem like a big deal.

Satoshi Nakamoto

To detect and reject a double spending event in a timely manner, one must have most past transactions of the coins in the transaction, which,   naively implemented, requires each peer to have most past transactions, or most past transactions that occurred recently. If hundreds of millions of people are doing transactions, that is a lot of bandwidth - each must know all, or a substantial part thereof.
https://www.metzdowd.com/pipermail/cryptography/2008-November/014814.html

Maybe, he did have plans in the future to scale it, maybe he had ideas, before his sudden disappearance.

I do think the initial genesis message probably suggests he was looking for an alternative to the current banking system, and maybe its already been achieved. Although, I think that Satoshi came up with a foundation of the solution, rather than the solution. I'm not convinced that Bitcoin is the solution to the whole problem with fiat currencies, just because of some of the limitations that have been brought up.