longfingernails It's the difference between the D'Hondt method of PR which was in place in the European elections and the UK's first past the post system for Westminster. It's also due to the fact that people do vote differently in European and Westminster elections. In the 2009 European elections, UKIP's vote share was 16.5%. The following year in the General Election they polled 3.17%.
The analysis of 2010 indicates that it was the presence of UKIP in key target seats that cost the Conservatives an overall majority. Labour have a built in electoral advantage under the current parliamentary boundaries that the proposed boundary changes (scuppered when the LibDems had a hissy fit over not winning the referendum on PR) sought to equalise. Which is always going to be difficult because as this article explains, even though there are regular boundary reviews, the data on which they are based is already many years out of date. The current boundaries were based on data from 2000 so will be 15 years out of date by the next general election.
Add to that, as also explained in the UK Polling Report article, the fact that as a general trend Labour seats are smaller than Conservative ones, it means that Labour need a lower percentage of overall vote share to gain substantial majorities. In 2005, Labour had a 3 point lead over the Conservatives and won with a majority of 60 MPs, whereas in 2010 the Conservatives had a 7 point lead over Labour and failed to win an overall majority.
I will try and find a different swing calculator, the one on the UK Polling Report site only allows you to include the figures for the main three parties. However, if you put the vote share from the European elections into their calculator (Labour 25.4, Conservative 23.93, LibDem 6.87 as per the BBC website then Labour end up with 319 seats, 7 short of an overall majority.
Now, these predictions are flawed in that they tend to work on the basis of a uniform national swing, which we don't see as much these days (it is still a factor, but maybe not as much as it once was). However, this General Election prediction, based on polls conducted between 5th April and 1st May 2014, currently gives Labour a majority of 40 MPs, with UKIP predicted to get a 14.44% share of the vote, substantially up from the 3.17% in 2010 but still without giving them a single MP.
Sorry, that took me a while!