Have you looked inside the spectrograph? Given both frames are giving weird results, it makes me wonder if something has been knocked out of alignment.
Another clue could be that bright line in the flat, and the odd short offset calibration lines, both appear at the same vertical offset. That could indicate something seriously out of alignment or light leaking in from outside.
It is probably still useable, with a dab of black paint over the scratch or just avoid putting the spectrum in that region.
Slits wider than the normal seeing are good for extended objects (eg comets nebulae etc) allowing more signal to be collected and for absolute flux measurements ensuring all the flux is collected (both with a reduction in resolution of course)