What are their potential applications and how is that a factor of the predicted demand for them?
You should ask that question for each one.
Eg. "Tugcoin.com". While it's a .com, which is infinitely better than .so, you have to ask yourself "Can I envisage a cryptocurrency that will be called Tugcoin, and will it be successful?".
If the answer is "hell's yea" then you have yourself a keeper, friend.
If you think "wait, there's never ever going to be a cryptocurrency called Tugcoin that will become successful" then you're wasting your time, resources and mental processing power by devoting those towards figuring out how to monetise that domain name.
As a concrete finishing example. If you had a domain like buycrypto.com, you could reasonably predict that more than 100 people globally would have an interest in buying that domain for a reasonably significant price, potentially even quite a high price. Therefore, your time, resources and mental processing power would be well spent seeking a mode of monetisation as you would likely be able to secure renumeration at a rate that would be significantly higher than the the combination of the cost of your time, resources and mental processing power.
So I think that's your first port of call.