As long as you see it as just a payment network, or a financial protocol you will have trouble seeing it's true potential. It is a protocol that enables digital scarcity in a decentralized manner without the need for trust. It allows for hard promises. After a few blocks a transaction is effectively irreversible. So it has to be extremely expensive to 'overwrite history'. I mean overwhelmingly expensive. It's a feature, not a bug.
Will this scale to the entire population. Of course not! Nobody is saying it will not change and evolve. The current implementation is great for a settlement layer while the LN would be used as a payment network.
We have been running out of IPv4 addresses for the last 30+ years, but things still work. People have been hard at work and incrementally IP was able to handle more and more devices (Address Classes, Classless Inter-Domain Routing, NAT, ...).
You praise the blockchain technology in general

Bitcoin doesn't equal the blockchain technology. I see that the blockchain technology isn't dependent on bitcoin and bitcoin value isn't guaranteed by the blockchain technology.
I think that you are bringing forward first mover advantage with the e-mail comparison. That the e-mail protocols mostly remained the same and this will probably make bitcoin remain dominant also.
I think e-mail can't be compared because of the simple nature of the problem they were solving. You can't exactly re-invent something as simplistic as a wheel. This new field of cryptos or financial protocols are a lot more complex and can improved a lot by re-inventive improvement. We have seen radical development with new generation cryptos in terms of speed, costs and practical usability. When comparing software, then cryptos could be compared better to the development of operation systems and other systems as complex. Not vintage communication protocols
