I would be emphasizing the difference in workflow for a discussion in either medium.
There is a fundamental difference between emails that do not require any further interaction other than reading the email - or deciding not to read it because of the title and who it came from - and emails that lead to further interaction.
As soon as you start a discussion with email, the interface becomes very cumbersome:
- Discourse retains the benefits of asynchronous communications while allowing the appropriation of many benefits from more synchronous/immediate communications such as text chat.
- Discussion continuity in emails is assisted by appending the contents of previous emails to the current response. But this is a massive duplication of data. I don’t mean massive in the sense of very large, but the frequent accumulation of small duplications which is rather like death by a thousand cuts.
- Email doesn’t really work if there are more than two people talking at the same time. So collaboration is restricted by the limitation of the email interface. Think about this example: when you have more than two people in the conversation, how do you work out what each person is responding to? You normally have to read the start of the appended message to see where they were in the conversation - most of us do this without even thinking about it but it remains another small cut. I can think of many email discussions where confusion has reigned when any of us assumed that others were a) up-to-the-moment and b) had read all other responses.
- Email doesn’t have features to handle mutating discussions. When there are changes in context, content, subject or recipients there is no easy way to see what has happened. Each change creates a disjunction. For example, a change in topic should be signalled by a change in the email subject line. But if that happens then most email clients, which only allow an email to be in one conversation, will create a new “conversation” using that new subject.