That is absolutely our intention! While users who run nodes can choose to forward content to anywhere on the Internet, we want to provide people a feeling of control over the usage of their bandwidth, as we believe that (along with being paid to forward traffic!) will increase the number of people who are willing to provide bandwidth to the service. Being able to say "I am only willing to help users get to Wikipedia, or the New York Times" is something that we think is very important to being able to have enough people using the service to provide the levels of security needed for everyone on the network.
> While users who run nodes can choose to forward content to anywhere on the Internet, we want to provide people a feeling of control ...
Are you concerned that,
1. If node operators have the power to control content, they will be held responsible for the content by authorities? For example, consider a node operator in mainland China who allows access to criticism of President Xi.
2. Unpopular content will be censored by node operators just like it is now by more centralized powers?
As these are more answers of opinion rather than factual questions about the network, I'm going to add a massive disclaimer here that "opinions may differ about what is or is not concerning".
1) No, because I would personally advise people in China do not whitelist sites whose purpose is to access such content. It is not even clear to me that people in China would want to run an exit node at all. There are billions of people in the world in countries other than China that can run exit nodes for that content so people in China will be able to access it without fear, and without anyone in China taking on that level of liability.
2) No, but with the great benefit that because everyone is using the same network, even if somehow (and I do not think this will be the case) there are only a small number of nodes willing to access that content (and one would expect that, in the eventual limit, this number should not be smaller than the number of Tor exit nodes that currently exist; though again: I think it will be much larger), then they will be cloaked within the content everyone else is accessing.
Essentially, it doesn't work to say "people who need access to a secure messaging service should use a secure protocol and everyone else can use Snapchat", and it also doesn't work to say "people who need to access criticism of President Xi can use Tor and everyone else can use the regular Internet". I want a billion people who are also doing things that would be considered insane on Tor--like browsing Netflix!--to be using this service, and to pull that off we need lots of exit nodes, which in turn means incentivized traffic and whitelists.
Put another way: I don't care if there are only 0.0001% of nodes that let you access criticism of President Xi if that is 0.0001% of some insanely large number and the result is "at least as many nodes as Tor", as that itself solves the #1 problem using Tor: that even accessing the Tor website already marks you as someone who is suspicious (as it is here in the United States and the surveillance programs we see).
There are only two options. One is all or nothing like tor: if you become a node you have to pass through everything. The other option is node operators can decide what they pass through.
With the tor model you get very few nodes, because most potential operators are afraid of being arrested and/or object to a lot of content. You don't get remotely as many nodes as you need to make the internet free and more safe.
With selection, you get many more nodes and, as saurik points out below, you still will have a few for material that some governments censor. So it seems to me a model with selection is the way to go.
