Questions like this expose the problems with Stevens' typology of variables. Some variables just aren't any of his categories; or are hard to categorize. Age is ratio level data in the sense that someone who is (say) 40 years old is twice as old as someone who is 20 years old. You just haven't got it measured precisely. But so? What difference does that make?
Dichotomies - here it really doesn't matter HOW you treat them for most purposes. E.g. in regression, you can consider them to be nominal, ordinal or interval and you will get the same results.
I wrote about other problems with this typology on my blog