Ok, I lied about starting a new thread.
This article is quite interesting I think - it says some of the same things as the ukpolling paper linked to further up, but crucially puts the same issues into IMO a broader perspective. Such as, the timespan between Boundary Commission reviews was reduced 50 years ago because of the instability factor of frequent reviews; such as, there has always been an electoral bias, but it has shifted from favouring Conservatives to favouring Labour in recent years; such as, the differential turnout issue has been increased because of the parties' logical (under FPTP) policy of concentrating effort on getting out the vote in marginal constituencies.
I do think that the ukpolling article that you linked to has a bias of its own, longfingernails (I would, since I disagree with the bias! - anyway, everything has a bias, so it's unfair of me to say that like it's unusual.) Firstly, it stated at the top as undisputed fact that the current system favoured Labour, and then gave several headings, only 1 or at most 2 of which actually did favour Labour. Secondly, it stated that bias as though it had always existed, which is not the case. Thirdly, it used biased examples, e.g. in the differential turnout paragraph it gave an example turnout of 90% for the Conservative constituency versus 10% for the Labour constituency. In this EXTREMELY pro-labour biased site, the respective turnout is given as 68.3% in seats won by the Conservatives, 61.2% in seats won by Labour.
End result is, however, that since 1997 the system has definitely resulted in a bias towards Labour. Just nothing like as egregious or as deliberate a bias as I have seen mentioned in some commentariat pieces.
I'd agree with you that FPTP is apt to spin a lead into a crushing advantage over third parties. Which is why I am thinking about voting for AV.