The qualitative reason is in the case of the pipe, the walls can be assumed for practical purposes to be rigid (i.e. they don't vibrate), and the resonant frequency of the vibrations in the air inside is determined by the boundary conditions. In other words, the shorter the air column in the pipe (more water), the shorter the wavelengths of the acoustic modes, or the higher the frequency. Or less water, lower frequency.
Whereas for the wineglass, the walls of the glass are thin enough that they cannot be assumed rigid - indeed they vibrate, and it is the vibrations of the glass that determine its resonant frequency.
With no water in the wineglass, the walls are not mass loaded beyond the mass loading due to the air inside. But as you add water, the mass of the water mass loads the vibration of the walls. Just as with a simple harmonic oscillator, the larger the mass, the lower the frequency of oscillation.