The plus 3 for the Tories in both polls is consistent. So too is the minus 3 for Labour as a result.
The significant thing isn't so much the number but that it's a high (wrong word but it'll have to do) enough number to probably be more than a statistical blip. Tempting to say "wait and see" for other polls, but that's silly as most of the other companies just use ICM's data and mess around with it (slight exaggeration).
Thing to remember about YouGov (and I
think this is why they produce weird results more often than the rest) is that they don't really measure public opinion, but the opinion of a fundamentally unrepresentative (and not just in terms of demography, but in terms of interest) section of the public. O/c they then do all sorts of things to make the final numbers look more realistic, but I think the unrepresentative nature of the people they ask sometimes slips through anyway, especially if opinion (and/or interest level) amongst the people in their panel about something is very strong, or if the people from the underepresented groups are politically (or elseways) unrepresentative. O/c they'll never have to worry about fixing that until they get a General Election badly wrong.
Eh. I know what I meant to say anyway.