Here's what was actually built:
Note that the W9CF Yagi java applet modelling program does not design or optimize the antenna. It just calculates what is possible.
Note the calculated gain is 15.53dBi for a 15 element yagi. This is best case as reality is usually somewhat less. Incidentally, I don't see a VSWR vs frequency plot on this design. There's more to making an antenna work than just gain.
That's normal for a yagi that tries to get a 50 or 75 ohm match from a folded dipole driven element without an additional matching xfomer. The first director usually ends up fairly close to the driven element. The 2nd and 3rd are also closely spaced. After those, the element spacing is almost equally spaced. No waviness anywhere that I can see.
If you look the the examples, there's the NBS 15 element 2 meter (144 mhz) Yagi, which has "almost" equally spaced directors.