I don’t think it will work well to provide the user with a description of each upcoming feature and ask them to rate or sort them. That requires users to try to imagine what the upcoming feature would be like in their life once it’s implemented. That’s much harder than a usability test, where users can try out exactly how a feature is implemented and give you feedback. For upcoming features, you can’t control what the users are going to imagine (if they can imagine it), so the results tend to be unreliable. You can try to explain the feature in detail (verbally, in text, or graphically) but that takes a long time, and it requires that the user hold a lot in his/her head when trying to do the ratings.
Instead, derive a set of questions about the means and ends of each feature from the user’s perspective. For each feature, you want to find out
The frequency that users have the goal that the feature achieves for them.
The importance of achieving the goal.
The degree the users have the resources to use the feature as it’s planned.
The degree the feature as planned may annoy or frustrate the user.
Combine these and you have the value of the feature (and priority, all other things being equal).
For example, say one feature for your social site is to send the user a text message whenever someone posts a response to a user posting. The goal of the feature is to eliminate the need for the users to constantly check back on the social site to see if there’s response. So the specify questions become something like:
How often do users check back at social site just to see if there’re responses?
How often do users find themselves wishing they knew if they had response yet?
How often do they have their phone on them when they’re wishing such things?
How disruptive or annoying is it to get and check text messages?
You may need more than four questions.
I’d take advantage of the usability test to do a mini contextual inquiry to answer these questions (or maybe “semi-contextual” inquiry if it’s in a usability lab rather than on the user’s home turf). Close the prototype and open the service as it exists now (or let the user open the competitor's service they’re currently using), and engage the user in a brief open-ended discussion on how they are currently using the service in order to answer the above questions. At the end of the discussion, try to get the user to put numbers down for each of the specific questions so you can compare the results of the upcoming features against each other. The hard part may be coming up with specific but comparable questions for each feature.