I'm not sure it's reasonable to assume that the distribution is the same between recovered and unrecovered children. Specifically because it ignores the circumstances observed by the reporting agency, and secondly because we tend to know where the non-custodial parent is.
https://www.ojp.gov/pdffiles1/nij/grants/304794.pdf
I lost my initial source, but this one serves just as well. I had initially been looking at data from the doj and ncmec.
Based on the NISMART-4 LES-SK, the number of stereotypically kidnapped children in the year from January 1, 2019, to December 31, 2019, likely fell between 52 and 306.

The Missing Persons reports from the NCIC show stranger kidnappings trending from 384 to 322 in 2011 to 2019.

This is from page 27-29 where discuss conclusions, statistics and data source quality (tldr: no significant change)
So, again: I got them by doing a cursory search and then looking at criminal justice statistics.
I'm not sure this has anything to do with the initial thing we were talking about anymore?