Well, no, that is how an antenna works. Or at least some designs like simple monopoles and dipoles. Don't Dunning-Kruger yourself.
For a dipole antenna like the omni the length is proportional to the propagation wavelength of the radio signal. Even for a single band, the length of the antenna is a compromise since it has to operate over several different carrier frequencies; for example 2.4 GHz (2.4-2.4835 GHz) and 5 GHz (5.15-5.825 GHz). See
Vertical Antenna Calculator
The perfect antenna for 2.4GHz has a bit of reflection at 2.4835GHz or a bit of loss. The length is a compromise. Note that for the 5GHz band, the 1/2 wavelength is close to the 1/4 wavelength for 2.4GHz band but not exactly the same. If you break open the omni antenna there is *one* element in there and the length/width is picked as a compromise among the bands needed.
The Alfa is different. This video (although I don't agree with all that he says) might be informative:
Note how simple the internal structures are.