The interplay between the individual cognitive functions (recognition, memory, organisation, retrieval, etc) is fascinating. On this very narrow function of "how many numbers can you memorize", I do badly.
In my work, however, I frequently need to transfer large tables of numbers from one place to another - usually from a book to a computer - which involves temporarily memorizing the position in the table and memorizing one, two or a few values; typing these into a keyboard; returning my gaze to the correct location for more; repeating until complete. In some cases, like strict libraries, I have to transfer from the library copy to a paper notepad and take that away to work on. On any single aspect of this task (even finding the right book and page, let alone memorizing numbers and opening spreadsheets) I perform much, much more slowly than my colleagues. In real-world tests though, I do the entirety of the task far faster, far more accurately and for very much longer than anyone I ever worked with.