Some strange ideas there. I won't go into all of it, but monitoring your output waveform, whether a sinusoidal observation or a trapezoid, doesn't tell you much about your signal unless you're seriously non-linear...enough to observe on a 4-5" scope screen (it has to be pretty bad to be observed). Much, if not most, of the signals on the bands occupying more than necessary bandwidth on modulated modes suffer excessive intermodulation distortion, which cannot be observed by an analog detector and oscilloscope. You can look "perfect" on a scope (nice, squared up trapezoid pattern with sides so straight they'd make a ruler blush) and still have a horrible signal. IMD cannot be observed this way.