This is a very small piece of polling good news after a run of bad news, but take it with a hefty dose of salt. Although YouGov are unlike other polling firms in that they seem to structure and weight their Scottish subsamples correctly, the size of the subsample is just 180, which means the margin of error will be very large.
The likelihood, I'd have thought, is that Labour are still in the lead in Scotland, and the apparent SNP lead is just an illusion of the margin of error. However, the hope must be that it's a straw in the wind suggesting that the calling of the election has focussed minds and that some people who voted SNP in 2015, 2017 and 2019 have "gone home" as a result. Time will tell when we see some full-scale Scottish polls.
It would make such a psychological difference if the SNP could claw their way back to roughly level pegging and hold at least 20+ seats. The jury is out on whether that is a realistic goal or not.