I think you are referring to the "Base Inductance" value which does not account for conductor size, and is 155 nH in every case. The actual values, when corrected for conductor size, are:
21SWG= 0.81mm, L=100 nH
24SWG= 0.55mm, L=108 nH
30SWG= 0.31mm, L=119 nH
These numbers are for an operating frequency of 0 Hz (DC). Therefore, there will be a very small change in inductance at higher frequencies.
Edit:
Just had a look at:
https://www.daycounter.com/Calculators/Air-Core-Inductor-Calculator.phtml
and I see that it also uses Wheeler's long coil formula, which is fine as long as you're aware of it's limitations (and most people aren't), but when you have the processing power of a modern computer available, why not use a better formula that's accurate for any coil geometry?