Some reviewers have said "this thing sucks, it doesn't work at all". You found that it sometimes works, but sometimes not.
I'd guess there is some factor "X" at play regarding the conditions in which it is used, maybe alignment of the wire under test or something?
A puzzle for sure. On the other hand, if a test instrument is not reliable, then how can it be trusted?
We have seen other instruments with design problems that were not discovered for several years, yet people trusted the results.
Watching the video review, we see a Klein Tools, PN# 69409 splitter being used to attempt to make the measurement. Looking at the quick reference, the first thing it the overview, they talk about 1 wire only and the wire must contact the bottom of the fork. There are also pictures that make it easy to understand. Of course, in this case an understanding of the splitter was required as well. We really have no way of knowing the specifics of people providing feedback. The best I can offer is to just show you the tests I have ran.
When I took the T6 with me for the day, I didn't have any problems if I could get to the wire I wanted to measure. However, all of those signals I would say were common and clean.
If I were trying to make these sort of measurements often, as say some sort of process control, I can see it saving time and being safer to use. Would I use it to make basic voltage measurements and trust the results, based on my experience with it, I would say in many cases, yes.
When I run my own tests, I am looking for problems. This is why I use an arb rather than a wall socket to test the meters. Obviously, I discovered cases where it can't detect the voltage depending on the shape of the signal. This is really what interests me.