For the reasons given above I find that highly unlikely. Proof of work and distributed consensus are at odds with the requirements of a global real-time processing network.
Once again, whatever people's views on bitcoin at least realize that mining hashes don't dictate how many transactions can be processed by a miner. They mine a block, then they include as many transactions as they think they can get away with and keep the latency low enough to propagate. Right now they are capped by a block size limit which is the real next hurdle. Some people think there should be no block size limit - latency already becomes a limit in itself.
To hammer it home - someone using bitcoin does not need to use any more electricity than a computer and internet connection. The miners use the electricity and the amount of transactions they process is not directly linked in any way.
I agree bitcoin does have its flaws, economically bitcoin behaves like a commodity, which might hinder consumer adoption. There are ways to get around that, you could potentially issue a anonymous currency of sorts backed by bitcoin, with a fixed exchange change rate. So, the prices could remain relatively stable. May be that could drive consumer adoption.
TBH, I don't know what bitcoin is going to look like in the next 10 years. It's like the internet all over again. Just like the internet, bitcoin value doesn't necessarily come from the underlying utility as a transaction platform or in case of the internet TCP/IP, but from its distributed , permissionless and decentralized nature of governance. We know for a fact, that's what drives innovation.
I was under the impression that more than 50% of the mining power is controlled by a very small set of Chinese miners. Could they not wreak major havoc by deciding to tweak the protocol themselves or not adopt certain features?
Similarly, what happens if the few bitcoin core developers decided to do something nefarious with the client. Surely they would be caught right away, but then how would the community decide on what software/protocol to use?
There is no doubt though that cryptocurrencies are potentially fragile. I do think though that pandora's box has been opened. Maybe bitcoin will stand the test of time like jpeg, C, C++, http, TCP/UDP/IP, zip, bittorrent, fortran, x86, SMTP, unix, RTSP, vi, javascript, etc that have been good enough and persisted with a huge network effect. Or it could end up being a jumping off point for something that smooths out the rough edges needed for mass adoption.
They don't want a federated pseudo-anonymous crypto-currency, just anonymous enough for theft without consequences, which requires them to hoard tokens on their home computer or a trusted server with extraordinary security measures as if they were gold coins and then pay a variable fee and wait a variable length of time (10 minutes!) to actually verify a transaction. I suppose gold bugs might be happy with it as a replacement for gold or stacks of cash under the mattress but as a currency it has glaring shortcomings. I don't think many people hate bitcoin, they just find the hype for it and its blockchain overblown and myopic. It's an interesting experiment, but it's unlikely to replace the currencies we have, I suspect that will happen at the transaction layer, not the currency layer (the underlying currency to a transaction is in many ways uninteresting).
The financial industry may well be ripe for disruption (indeed some startups in the UK have started), but the founding tenets of bitcoin are such that it is not a suitable replacement.