Exporting to WAV should result in exactly the same bytes, no matter what app is involved, because the WAV-format is a well-defined fixed format. Of course aside from differences in resolution-settings like 16- or 24-bits, or nr of channels.
Like MisterX said, playback is another thing, and like LPChip said, recording is another thing. Then the soundcard is involved in the process. Clockspeed, AD/DA-conversions, nr of bits, buffers.
This topic made me focus a bit more on the 'save as Wav' option, checking some details.
I already often used the nice option to save only a selection of the pattern-row, or a single pattern, as a WAV.
So far, if I wanted to make a WAV-file of only channels 5 and 6 of the song, I muted the other channels.
Now, to my pleasant surprise, I discovered the 'Channel mode' option. I tested with a 12-ch XM-module, and it resulted in 12 different WAV-files, numbered. Very useful. If the export mode is stereo, then the embedded panningcodes in the selected channels remain valid, if the export mode is mono, then of course not.
Since I make most of my pieces with four discrete output channels, I checked the 'save as quad-Wav' as well. This brought up some complicated issues on its own, so I will bring it on, but in another topic.