On this video they created an UVC lamp with a 400W MV tube and they drove it with what I think is a 60W incandescent bulb, which seems to be underdriving the MV lamp. But the interesting thing is to see it sparking, I've never seen an MV sparking when warming up.

I think "flickering" is a better word than sparking. I'd wager that since it is underdriving the lamp, the cathodes don't get hot enough to be reliably emissive, and the arc hunts around as conductivity changes.

I've tried this using a broken lamp before, and it really doesn't work. The arc tube is intensely cooled by ambient air so that it can never run up, even on a full power ballast. The UV light production is quite poor as well - not sure why though.