Imports have actually spiked on the rumors even though it’s Sunday: normally we import 400-600 posts an hour from Tumblr, last hour it was over 72,000.
I'm sure that a lot of the stories about this will make a big deal out of the spike, but it's not all that large a figure when considered either in terms of Tumblr users or of the number of posts that service currently hosts.
A brief check of the Tumblr about page will show that Tumblr currently supports over 108 million different blogs that contain over 50 billion blog posts:
Assuming the exodus remains constant at 72 thousand posts an hour, it will take close to 707 thousand hours (or over 29 thousand days) to transfer all the content off of Tumblr.
Suddenly that number is not so impressive, is it?
My point is that we have yet to see a serious negative reaction from more than a tiny vocal minority. While I expect the rate of exodus to increase, in absolute terms it's really not all that amazing. And even if we assume that there are other Tumblr users fleeing to other services, we're still not seeing a huge number of users abandoning the service.
If I were Tumblr I would not be concerned about the exodus until it reaches and holds at at least million posts an hour. That is still only a tithe of Tumblr users but it is also a significant number of the most active and most vocal members. A good part of the value of Tumblr is the community of users (and not just their content). The most active users are the linchpin of the community, and if they go elsewhere the community could wither on the vine.