>>As a simple example, how many people would use a tumbler if that started being seen as probable cause for a money laundering investigation?
Filesharing is arguably in this category and maybe shows that it's possible that something not generally used for legal purposes and not in the mainstream can nonetheless become widely used. But yes there is certainly a risk.
>Some people value that but it seems unlikely that it's enough to make up for the smaller network, processing delays, and exchange fees.
Right now cryptocurrency is in its infancy, so it has a lot of drawbacks, but it has obvious advantages that anyone can see, like being able to send money from your computer, without first registering with Google Wallet, Square, Venmo, etc, to anyone, anywhere in the world who also has the software installed on an internet connected device. This is a fundamental innovation in money transfer. You're right that it may not be enough to overcome the advantages that large trusted third parties can bring to make it go mainstream, but the advantages it has are obvious enough that people can see it potentially doing so.
These things are always hard to judge but there's certainly plenty of prior art: there's a ton of CS history for distributed consensus, distributed attestation, etc. prior to 2007 and the proof of work concept goes back to at least the 90s – see e.g. http://www.hashcash.org or http://www.hashcash.org/papers/bread-pudding.pdf.
Anyone who spent time on cypherpunks-l in the 90s would also be familiar with the discussions of e-currencies like David Chaum's 1989 digicash (https://en.wikipedia.org/wiki/DigiCash) which did not use proof-of-work but did cover a lot of the same ground for anonymity, etc.
Of course Bitcoin built on previous ideas, but the ideas were more theoretical, in the realm of academia and obscure cypherpunk mailing lists.
The components of an iPhone (e.g. the touchscreen, microprocessor, RAM, SSD drive) OTOH were already mass-produced and widely adopted in consumer electronic.
I don't want to understate the iPhone's impact. It spurred the development of better LCDs, touch interfaces and especially SSDs, but it was undoubtedly still working off a more established base than Bitcoin.
The consensus algorithms predating Bitcoin required some percentage of participants to be trusted identities, so were quite different than what Bitcoin pioneered.
Hashcash was a very limited experiment that had no consensus system. Utilizing the proof of work idea of Hashcash to create an identity-less consensus system was more conceptually groundbreaking than iPhone's marriage of touch screen mobility and personal computing.
While the iPhone added a personal computer to a mobile phone, in a user friendly touchscreen package, the blockchain created something that was totally new: distributed consensus without known and trusted parties.
Filesharing is arguably in this category and maybe shows that it's possible that something not generally used for legal purposes and not in the mainstream can nonetheless become widely used. But yes there is certainly a risk.
>Some people value that but it seems unlikely that it's enough to make up for the smaller network, processing delays, and exchange fees.
Right now cryptocurrency is in its infancy, so it has a lot of drawbacks, but it has obvious advantages that anyone can see, like being able to send money from your computer, without first registering with Google Wallet, Square, Venmo, etc, to anyone, anywhere in the world who also has the software installed on an internet connected device. This is a fundamental innovation in money transfer. You're right that it may not be enough to overcome the advantages that large trusted third parties can bring to make it go mainstream, but the advantages it has are obvious enough that people can see it potentially doing so.