Decisive majority for independence in sensational post-referendum Survation poll

Should Scotland be an independent country? (Survation)

Yes 54% (+6)
No 46% (-6)

The percentage changes are from a Survation poll in April, which as far as I can see is the last time the firm polled on independence.

Although these figures are only a little better for Yes than we saw in the Panelbase poll last night, they're still significant because they substantially increase the chances that the pro-independence majority is really there.  The Yes lead in Panelbase was within the margin of error, but tonight's lead is not.

This is by some distance the best ever showing for Yes in a Survation poll - the previous high watermark was 51%, which was in the only previous poll from the firm to ever show an outright Yes lead (in March of last year).  There was one other which reported a dead heat, but other than that it's been wall-to-wall No leads.  Although there was a spell during the independence referendum when we regarded Survation as one of the three relatively Yes-friendly firms (alongside ICM and Panelbase), that changed radically as polling day approached, and Yes found it frustratingly difficult to make the sort of breakthrough in Survation polls that they were managing in polls from other firms.  48% was the best they ever did with Survation prior to the referendum, even at a time when ICM were putting them as high as 54%, YouGov were putting them as high as 51%, and TNS were putting them as high as 50%.  And remember that it's actually got harder for Yes to do well in polls since the referendum, because weighting by recalled referendum vote was introduced to correct a small systemic error.  So basically what I'm saying is that 54% in a Survation poll is a really big deal, even if some people might have been dreaming of something even better.

UPDATE : Much of what I wrote above was based on the assumption that this was an online poll, in line with the vast majority of past Survation polls on independence - in fact, it was a telephone poll, so the direct comparison doesn't work quite so well.  (Although it is directly comparable with the private Survation telephone poll for Better Together on the weekend prior to the indyref, which was only published because - like Survation online polls at the time - it was relatively unfavourable for Yes.)  And as Calum Findlay notes in the comment section below, Survation seem to be saying that this poll wasn't weighted by recalled referendum vote, because they found evidence of false recall - but they don't reveal in which direction.  Could that mean past vote weighting has slightly distorted the results of past polls?

UPDATE II : From the datasets, it's clear that Survation think that many people are wrongly recalling (or claiming to recall) that they voted Yes in 2014 when they actually voted No.  If Survation are wrong about that, this poll may have overestimated Yes.  But if they're right, it's possible that many previous polls since the indyref have underestimated Yes.  A third possibility is that false recall may have only emerged as a factor due to the outcome on Thursday, with some people now embarrassed to admit that they voted to keep Scotland in the UK.

Survation also seem to be assuming quite considerable false recall or embarrassment in relation to EU referendum vote, because the weighted figures have Remain winning in Scotland by 73% to 27% - way above the real figures of 62% to 38%.