I'm not sure the latest Survation MRP update is as significant as people are assuming, because if the fieldwork dates are anything to go by, it's based on about 80% of the same interviews as the previous update. The top-up of new data hasn't made the situation better for the SNP but it hasn't made it dramatically worse either. And as an anonymous commenter pointed out on the previous thread, the credibility of the projection that the SNP will get only 10 seats hinges on whether you believe the popular vote figures, which show Labour 11 points ahead of the SNP. That's possible, but it's significantly out of step with the conventional polling (so far). If you think the SNP are a bit closer to Labour than that, it would mean they'd probably end up with more than 10 seats.
The most recent batch of conventional polls show...