Fantastic questions.

1. I know everyone has had this first-year-econ bong-rip moment already, but of course all money is ultimately a reflection of confidence: confidence, above all, that it will pass -- that someone else will accept it from you, as you accepted it yourself. (The great term of art for anything that functions as money is that it's "passing current".) Usually that confidence stems from long cultural history (precious metals, cowrie shells) or state authority (legal tender for all debts public and private). The rough deal a crypto faces is what anyone novel currency faces: how do you get people to take it? How do you get it to circulate? One answer is, you make it possible to do something with it you can't do another way (in the case of Bitcoin, and I'm not being sarcastic here, it was buying good drugs online). The problem arises if the answer is, promising people it will be worth more tomorrow than it was yesterday. That means cryptos are uniquely vulnerable to being turned into Ponzi schemes: if that's the best way to drive adoption, and driving adoption is how you get rich, then you keep constructing new ways to bring in fresh suckers to prop of the value of current holders. And like all Ponzi schemes it works great until it doesn't. So cryptos are not Ponzi schemes inherently, but in the absence of any other value proposition they tend to turn into them.
2. I think a lot of academics (myself included) are very skeptical about the current crypto landscape addressing any major issues that it was originally intended to -- in part because the technology and culture of crypto have evolved from their early days, moving from a post-national currency for radical libertarians to opt out of statist banking systems, into the jankiest of the asset classes for financial speculation. An abstract financial vehicle, wobbly as any other, but where the guys in the room wear hoodies rather than polo shirts. They have become, in many ways, solutions in search of problems to solve.
3. Crypto in general is, by design, phenomenally inefficient. It's meant to be a giant resource hog, an extractive industry, like strip mining. That's now changing (Ethereum is a good example of this), in part b/c of changing attitudes: the population that still thinks of crypto as a digital version of precious metals is small, and shrinking.