Why do you think measuring equipment and safety ratings were invented in the first place? Because better men and women than us paid the price for not having them. I don't think anyone expects a hobbyist to work on industrial equipment. However, the gear should be safe with anything commonly and reasonably found inside the home. That includes poking the mains with a meter apparently rated to do so.
I have actually seen the results of using a poor defenseless Simpson 260 on 480VAC with the switch and leads set to either Rx1 or some current range. It was hard to tell since the meter approximated a hand grenade. The tech (an HVAC guy) decided to stay out of the electrical end of the business, like forever. There was a bit of plastic embedded in his chest and the engineer with him needed a change of pants. That's the problem with working out of your job classification. 480VAC with 100k Amps of available fault current is no place to mess around.
Fortunately, there was no plasma arc between the phases. That would have been exciting!
I don't believe there is an appreciable difference between probing the residential panel and a wall outlet 30 feet away. More often than not, the available fault current is on the order of 5000A and, possibly, but rarely, as high as 10,000A. These are not insignificant numbers! But the CAT ratings are for impulses - lightning being the most likely (since we are not running street cars in our houses).
In terms of residential troubleshooting, I don't think I would go out of my way to have a CAT III rating although just about every meter claims to have one. And it's all CLAIMS because, AFAIK, UL listings are fairly scarce. There are a lot of CE marks but those are self-certifying. Maybe true, maybe not... How many of the Chinese meters carry a UL Listing? That Uni-T doesn't AFAICT and the Aneng certainly doesn't. Both are CE marked and both carry similar ratings with the Aneng able to withstand a slightly higher voltage applied to the resistance scales.
http://content.fluke.com/promotions/promo-dmm/0518-dmm-campaign/dmm/fluke_dmm-chfr/files/safetyguidelines.pdfWe should also take into account that people are forgetful. Even if you know a meter is unsafe for anything but low voltage now, will you still know when you pick it up in 5 years, when the hobby has died off but the multimeter is still around? Will a random family member?
Any CAT II meter should be fully qualified for residential (family member) work. CAT II is so low on the totem pole that it is almost always a 'lesser included' with some CAT III rating. Neither are going to survive a lighting strike on the pole at the end of my driveway.
I'll concede that the ratings are probably bogus on any meter not labelled FLUKE. If the user measures voltage with the probes in the 10A jacks, things are going to go bad. The very best meters try to signal when the probes are misconnected. But, in the end, if the probes are set to measure current and the dial is set to measure current, the meter won't complain until you measure voltage! But this has nothing to do with CAT ratings. But it does go to fuse type and voltage rating. Will the fuse extinguish the arc or not? Wanna bet?
Really, the safe bet is to only recommend UL Listed CAT IV 1000V meters. I don't know if anybody makes one because all I see are CAT IV 600V but still... But it only counts if the meter is UL Listed or ETL Listed. CE Mark doesn't mean spit.
Another thought: Just because a person knows a little electronics and is fully qualified to blink an LED with an Arduino doesn't mean they should be working on mains. Clint Eastwood said it best: "A man has to know his limitations!".
The problem with multimeters is that they are so simple to operate. So simple, in fact, that many users don't stop to THINK. Meters should power up with THINK on the screen!
Not saying it won't happen tomorrow but in 60 years, I have yet to pop a fuse in a meter. Heck, they didn't even have fuses when I started.