I am also a fan of VTVMs. They have their place.
I have seen a lot of discussions, some turning into flame wars, on the net about analog vs. digital test equipment. I learned long ago that there are no perfect test instruments. You need to determine what works when, how you need things displayed and how much accuracy you really need, or makes sense. Though I have multiple midrange bench/handheld Fluke and HP DMMs on the bench, I restored a Heathkit IM-25 and it sits in the middle of the bench. Not actually a VTVM but a solid state version of one. It is accurate enough for many uses and I have lab calibrated instruments if I need higher accuracy.
What is really important, as pointed out below, is the large analog meter. For many measurements a moving needle meter is much easier to read and easier for my eyes to follow than a digital readout. Some meters have bar graphs but not much resolution compared to a large meter. Where VTVMs really shine is when tuning things where you are looking for a peak or null, or when trying to follow signals that vary all over the place. In many cases you don't even need to stare at the meter to get readings (like on a digital meter) but just keep an eye on it in your peripheral vision as you make some preliminary adjustments. A scope can be used in some cases but might be overkill. I also have an analog frequency meter for tuning and following frequency changes. I have spectrum analyzers ranging down to 5hz but an SA might again be overkill or too much setup.
I still have a bunch of Heathkit, RCA and EICO VTVMs and will use one from time to time, mostly on older tube equipment where I might worry about permanently damaging my Fluke meters.
Bottom line is to determine what you want to measure, how much accuracy is needed and how you want/need to see the display. VTVMs still have a niche to fill.