-
I'm not hugely in favor of list based PR on the whole, mainly as I like the idea that MP's are tied to places, it makes it harder for them to tell you to get knotted. Even region based D'Hondt isn't quite close enough for me for this, as the areas are too large. It also prevents independent candidates from standing.
I think that personally I'd lean towards larger constituencies, with a number of top-up MP's that are based on the percentage of votes cast (maybe using a regional D'Hondt system) much like the London Assembly is, but I'd be happy to go with something else if I could be persuaded of its merits. I'm also not sure if the top-ups should be on the basis of the initial vote, or a separate one, as I could be persuaded to vote for an excellent constituency MP, but not like their party. Also see the independent candidates point above (I think that I've just explained that I'd rather it was separate).
-
Even region based D'Hondt isn't quite close enough for me for this, as the areas are too large. It also prevents independent candidates from standing.
Assuming a constituency size equivalent to about 5-10 current constituencies, an independent could still become an MP by taking 10% of that region's vote. Not so useful for the extreme of local-issue cases, but I would imagine that independents who garner enough support to be competitive in one current constituency would also have relevance across a wider geographical area.
http://www.bbc.co.uk/news/election-2015-32601281
This gives the split under the D'Hondt PR system (which assigns seats using PR in regions, so Scotland is handily self contained). So the SNP would have got 31 seats rather than 56, just over half which isn't bad given their poll (thanks to your figures) of 49.97% of Scottish votes.
More opposition and a fragile majority.