And now we're saying the same thing. If you accept my premise that the exogenous testosterone dose sets free testosterone, then you can use any free T calculator to see that a lower SHBG requires lower total testosterone to maintain a particular free testosterone level.... All I'm saying is that something in low shbg seems to correlate with lower than expected t numbers so one naturally assumes it's associated with shbg in some form or another.
It's also plausible that low SHBG has a separate effect on underlying metabolism, in the fluid analogy effectively enlarging the hole in the lower tank more than predicted. This effect is not present in my data in the graph above, even with SHBG values ranging from mid-20s nMol/L to mid-40s. But if present maybe it's only significant at even lower levels?