(so after reading the tektroniks' pdf, sorry for the delay, i'm pretty busy with a proyect
)
so the only danger of floating a oscilloscope is that the chasis takes the voltage reference of wherever you plug the oscilloscope's gnd clip and the added parasitic capasitance and indutance, and the ringing caused by those two.
luckly, my dso has an all plastic case and handle with ruber buttoms, so no problem with the earth reference there, except for th usb, the bncs, and serial conectors in the back, there's commonly no metal at fingers' reach
(not meassng high voltage anyways)
so summarising...the only diference, in practice, between the transformer and the ch1-ch2 methods is the amount of probes you have to use.
the ch1-ch2 is cheaper but you have half the samples per second, and only one channel, using 2 probes
and floating the oscilloscope you have one and a half channel, with one with full sampling rate, but you get parasitic capasitance and inductance, and some dangers above 40V.
i got it right, didn't i?
just one more question, how much inductance and capasitance would that be? my scope says 13pf, but how much does the chasis add? so as to know how "dangerous" could it be to my circuits.
ps:but i still don't get the "...stressing the oscilloscope's power supply." part though