如何安排这些数据？

Although our actual requirement is slightly more complicated, I would really love to hear some opinions on this:

Lets say my datastore entities are like so:

type BlogCategory struct {
  Slug string

  Name string
}

type BlogPost struct {
  Slug string

  Title string
  Featured bool
  Tags []string
}

BlogPost have BlogCategory as their parent. So each BlogCategory has zero or more BlogPost.

Q: Now, I want to write a datastore query which returns all the featured blog posts, across all the categories.

I realise that I could have just made category a property in the BlogPost kind, but I wanted ancestor queries in the first place to get consistent results in my test cases.

Are there any more elegant ways of tackling this? I have a bunch of solutions in mind, but thought I would ask the hive mind before committing to one path.