Hi David,
Top ten in total is the main focus. However, this would in the long run mean that the big projects (the Wikipedias, Commons) would dominate every year, whereas something that would have a very big impact on a smaller project would never be in the top ten. In order to balance this, the Community Tech team plans to spend roughly 75% of its time on the top ten wishes, and 25% on wishes related to smaller groups ("This will include campaign and program organizers, GLAM participants, smaller projects like Wikisource and Wiktionary, and stewards and CheckUsers".) Additionally, while the Community Tech team is mainly doing this as a way to let the communities decide what they should work on, we saw in 2015 that it's been used by other developers. You can look through the status column on this page to see what's been happening with those requests:
If we were to pick the top of each category, I think we'd have had to think about the categories more in beforehand. As it is now, they're a tool to make it easier to get an overview of the proposals, and have actually changed during the survey as we've seen what's been popular and what hasn't.
TL;DR: Top ten, but yes, other wishes might be worth pursuing anyway, we're trying to see if we can do something about it.
//Johan Jönsson
--