In 1995, the Internet was a vast improvement for things like email and file transfer and online forums and some games. It delivered actual value for 30 million people. Yes, there was more that it could do. But it delivered value on a daily basis for a lot of people. (And that was also true in 1985, albeit for something like 30,000 people.)
Bitcoin will never offer lower fees and greater convenience than existing payment networks. If it actually becomes a threat, those existing payment networks will get their shit together. And they can be far better than blockchain-based alternatives, because blockchains impose unnecessary technical limitations.
Look, for example, at M-Pesa. It's an electronic currency that leapfrogged existing payment systems. It started about the same time as Bitcoin. Except that it's actually useful for normal transactions and has real usage. In 2016, it was averaging 190 transactions per second, about 100x what Bitcoin is doing. Transactions are instant, and its fees are already lower than credit cards.
We could have that today in Western countries, but we don't because incumbents don't want to change and we're mostly fine with that. But if incumbents come under threat, they can do at least that well, and probably better given their much larger tech and research budgets.
The lightning network is just getting started which (once mainstream) is going to dramatically lower fees and transaction speeds for a ton of bitcoin use cases.
And even if some of the most compelling use cases are niche- (collapsing economies, illegal activities) those alone will keep the currency value afloat and if eventually it matures and offers lower fees than credit card companies and developing nations where many people don’t have bank accounts can leap frog our legacy payment systems then I think the tech will be here to stay.