Before the point system was adopted, different sizes of type had individual names.
And the actual sizes of type bodies (at least) changed when the point system came in. For example, Small Pica was closer to 10½ points than to 11 points.
Where did these sizes come from? Were they just chosen at random to baffle the innocent?
One thing I thought earlier was that either Ferguson's system, where the different sizes of type were expressed as different numbers of lines in a constant space, or Bruce's system, where the difference from one size of type to the next, might have expressed the logic behind the old system. But, in fact, the dimensions of the old types didn't match either of those schemes exactly any more than they matched the point system.
Something I recently came across in Googling for information related to another thread has suggested another possibility. The type used in the 42-line Bible was trimmed by Gutenberg; originally, it was cast to make 40 lines on a page, but Gutenberg found that he would need to have a bit more economy of space in order to make the price point he needed to achieve.
Now, books are made from different sizes of paper, as well as being folded different numbers of time. So one has books described as being on pages with names like "Demy octavo" or "Royal quarto" or "Foolscap duodecimo".
So one could suspect that the six basic type sizes originally belonged to more than one series, each series following Ferguson-type logic, but with a different page size (less margins) as its basis. However, it is entirely likely that soon after, one printer just copied another, and this basis was forgotten, partly because customers would end up specifying mismatched type and page sizes.