At the risk of being shot to pieces here, but to my mind, as has been pointed out, voltages the accuracy at best is 3%, even the cheapest DMM's manage way better than that, but to me the real advantage of a scope is the visualization of what is going with the signal and in many service manuals, it shows at various test points on a device what the signal should appear as and that for me is priceless.
My favourite example for this is the following story that happened to me years ago:
I had just converted a studio tape recorder from mono hollow state to a home-made stereo circuit with semiconductors. The circuit was meant for opamps type 709 which by the time weren't state of the art anymore so I thought it a clever idea to use NE5532 instead.
Once everything was completed and mounted, next step was adjusting the playback gain and equalisation. Out came the really expensive reference tape for 15 ips (a few hundred deutschmarks at the time) with the recorder connected to a AF voltmeter. The first pass looked ok but there should have been some room for improvement. Rewind and next pass. Funny, though, the frequency response showed yet more roll-off at the upper end. Cranked up the EQ to compensate, rewound the tape, next pass. Still more roll-off at the upper end. By then, the EQ pots were at max. Lots of head-scratching...
That's when I decided to heat up the scope and things instantly became devastatingly clear. With the much faster opamps the whole playback amp was oscillating like mad. So much that it slightly erased my reference tape via the playback head - a little more with each pass.
A few picofarads across the opamps cured the oscillation and the EQ worked as intended. Just my expensive reference tape was ruined.
My AF voltmeter has an ouput for a scope and I've never again done anything without watching the signal on the screen as well.