Well, no matter to inscribe a circle into the square; the same circles abo­ve, resized to fit the square height show their Φ ratio at the height of the navel, with the upper circle's center to the point c (page 24), like the arms' radius ch! Here the navel seems slightly displaced (the handy cir­cle flattened, still under revision, from that navel might ­not match the right cir­cum­fer­ence); so it needs to be verified in al­ter­na­tive models.

After all, it may be hard to argue about a centimeter up or down, that can vary with races, general characters, age as by the weight of ab­do­men and skin; the same to exactly reset a distorted multiple geometry.
However, over a lot of hand made models that I could examine, the Mi­chel­an­ge­lo's David seems the most reliable, the entire ab­do­men looks more realistic; I added too the dummy ac­cu­rate­ly man­u­fac­tured for sci­en­tif­ic use in acupuncture, both meeting our target… (PDF 5.7Mb) but perhaps not yet enough, as we're going to verify later.