pfm wrote:
Many advocate the use of an isolating transformer.
The oscilloscope probe and ground
are then used like multimeter leads.

....which will get you exactly where you want to be.

Tektronix and others maintain this is a dangerous practice

....to the bottom line of Tektronix, Inc.

and may result in cumulative damage to the oscilloscope.

....of which they couldn't provide a convincing example.

Differential probes are one answer, though quite expensive.

See "bottom line", above.
The claim is bull****, just like MICROS~1 telling you
that software other than theirs will cause you problems.