A signal cannot propagate in parallel plate waveguide 1/2 wavelength wide with the E-field parallel to the plates. It is at cut-off.
With the E-field normal to the plates there is no cut-off for the fundamantal mode which is a TEM mode. With half-wavelength separation, no other mode can propagate.
One way to understand this is to think about what would happen to a plane wave if it arrived at the edge of a stack of metal plates.
If there was a plane wave, vertically polarised (electric field vertical), travelling in a horizontal direction, then any number of horizontal thin metal plates could be put in place and it would make no difference to the wave, whatever the separation. This would be a vertical polariser, for instance, and each gap is a parallel-plate waveguide in the fundamental mode m=0.
If the wave was horizontally polarized, then there would be a reflection from the metal edges, but if the plates are more than a half-wavelength apart some power will travel between the plates. The modes that would be excited would be the m=1 (more than 1/2 wavelength) m=3 (more than 3/2 wavelengths), m=5 (more than 5/2 wavelengths) etc. The modes for even m won't be excited unless the wave travels at an angle to the horizontal, and they will appear when the separation is more than 1 wavelength, 2 wavelengths, etc.
For the vertically polarised wave, if the wave is at an angle to the horizontal, other modes also appear, in the same order.