Here is documentary that might give you an impression how creating new local currencies can successfully solve real-world problems, blockchain technology fits very well into that because of it's decentralized and digital nature, making it a lot easier to create, distribute and convert currencies (although the documentary doesn't focus on that aspect, but is explained more in depth on the website).

And here some research:
https://www.grassrootseconomics.org/research
 

Ah, OK, so I think your point is more along the lines of "it's relatively irrelevant" which I really didn't even take into consideration given the already significant effects of crypto (for example in terms of CO2 emissions which are already about 0.1% of total emissions) and the extreme growth the market has seen and is seeing.  
I would say the percentage of transactions is not really the main factor here, more the wealth held in it, and crypto already is well on it's way of holding 1% of the world's wealth.

Maybe you don't believe it solves any genuine problem for you in particular or for society as whole, but creating money out of nothing clearly is a very powerful tool if you happen to wield it, so I don't see it going anywhere.
Even if it does end up just being used in the black market economy, that is estimated to be >10% of global GDP.

Thanks a lot for you take.

One thing I would note is that cryptocurrency as a cause area is independent of cryptocurrency having have a net benefit or a net harmful effect; potentially cryptocurrency could destabilize global financial systems, so if one has a less positive view on cryptocurrency, regulating cryptocurrency (whether by governments, or by self-regulation within the ecosystem) and making sure at least some cryptocurrencies have a positive impact (thus reducing the overall net harm) could still be a potential cause area.

To address a few other points:

Massive energy consumption. I'm thinking of bitcoin, the mining of which reportedly sucks up ~0.3% of the world's energy consumption. If so, that's a lot.

Indeed, bitcoin is outrageously bad in many aspects of its technology.
This isn't a universal problem for cryptocurrency though, the whole Nano network for example could approximately be powered by a wind turbine while having much higher throughput and speed compared to bitcoin.

It goes to show how the crypto market right now is not guided by the right values, the bar for positive change here is quite low.

This one is not strongly held: It may be a talent black hole. I don't know if this view is some vestige of at one time having more leftist sympathies, but I still worry that complex financial markets can siphon off scientists from having a much higher impact.

I am pretty sure that's a real problem, the complexity of blockchain technology is quite high, and there are powerful monetary incentives while the social benefit is far from being automatic.

General reservation: For comms reasons I think EAs should be conservative / cautious on the margin around controversial topics or cause areas that'd distort our image in such a way that could damage our LR growth.

I agree, it might not be suitable as a widely presentable cause area in the same way poverty is.

Funds criminal organizations. This view may be a little old fashioned, but criminal organizations have large negative externalities. Particularly Latin American drug cartels which may benefit from cryptocurrencies.

I don't think it's old fashioned but still highly relevant.
For example, cryptocurrency being banned instead of being regulated in a sensible way could potentially to lead to criminal organizations benefitting from cryptocurrency while devaluing legal currencies, so I see a real risk here.

Keeping the mostly-efficient market hypothesis as a prior, I'm skeptical of most propositions that include the first part of that quotation.

I am not sure the concept of an efficient market is a very useful concept with regard to cryptocurrencies as it's now.  
As a former poker player I see more similarity to poker in that there are rational ways to get an advantage, but the market as a whole is rather irrational, driven more by circumstantial information (eg Elon Musk making a tweet) rather than well-informed investors.

The big difference to poker is that cryptocurrency is not a zero-sum game for the players, and at least in the past it was a positive-sum game in a rather dramatic way, as it was such a relatively small market which had incredible unrealized growth potential.

This is decreasingly the case as the market becomes saturated, but it seems realistic the market cap of cryptocurrencies could still grow to a multiple of its size over the coming years, making the average crypto investment potentially quite profitable.

Unfortunately you seem to miss the point.

I am not here to say how great cryptocurrency is, although it can undoubtedly be extremely profitable (for better or worse), but to point out shaping it positively is important.

I guess your argument boils down to suppressing cryptocurrency being more viable than shaping its development positively, but history shows that many goods cannot be effectively suppressed and if people derive a benefit from it they will continue to use them, with all the disadvantage a black market brings (I am talking about the drug market).

Also you have to seem a rather idealistic conception of government, where even if people have to deal with the consequences of dysfunctional monetary system (eg Venezuela), it's still preferable for people to suffer the consequences than to provide an alternative that is not sanctioned by the state.

I agree that aggregating suffering of different people is problematic. By necessity, it happens on a rather abstract level, divorced from the experiential. I would say that can lead to a certain impersonal approach which ignores the immediate reality of the human condition. Certainly we should be aware of how we truly experience the world.

However I think here we transcend ethics. We can't hope to resolve deep issues of of suffering within ethics, because we are somewhat egocentric beings by nature. We see only through our eyes and feel our body. I don't see that ethics really can adress that level meaningfully, it requires us to abstract from that existential reality.

For me the alternative is a more pragmatic ethical framework. It acknowledges we are not just ethical beings, but that ethics is important on an interpersonal level. From that point of view helping more people can be the right thing because we are aware we generally cannot truly resolve others suffering on an individual basis. So we are in effect helping the greater system of society or humanity. In that case there's no problem helping a group instead of an individual. We are not trying to help "at the root" - which we may only be able to do for ourselves or perhaps people close to us - but contribute to society in a meaningful way. And on that level there's a practical difference between helping one person or many.

In practice, for me that means I do take effective altruism into account, but also acknowledge its limitations. I'd say everyone does that implicity or explicity.