Seems to me that your calculations on "super radars" needs to consider the number of four digit "prefixes" that are available to make them up from. 1111 isn't one of them, neither is 0000, if you see what I mean.
When I first read your post, Paul, I thought you were correct and my 1:100,000 figure had to have been incorrect. But it seems like I was right after all!
Paul, you are correct that there would have been no "super radars" starting with 0000 or 1111, for example. But, neither would there have been any regular notes starting with those digits either.
So, consider the 10,000,000 note series starting with prefix 5612. Any "super radars" would have to end with 2165. So we can construct a general case: 5612xxx2165.
Replace the "xxx" with all numbers from 000 to 999. 100 of those 1000 possibilities will be radars (101, 202, 303, 010, 888, etc).
So, there'll be exactly 100 "super radars" for that series of 10,000,000 notes, or a ratio of 1:100,000. Repeat for all the different possibilities for 4-digit prefixes in the 1979 series, and you still get the same odds, 1:100,000.
My brain works much better in the late afternoon than it does first thing in the morning.