When you say "get another chance," you make it sound as if they get two chances. But that's not true, as this article explains:
this article explains.
"It is a mistake to regard this as voting a second time, and a mistake to say that this preference is worth more than anyone else's preference. At this stage no votes can been deemed to have been cast, except in the sense that voters have completed their ballot papers. Examining the preferences is a preliminary process to the final casting of the votes. The first preferences of voters for the bottom candidate are disregarded, not cast. They play no part in the final decision."
I understand that you're saying that the very fact they're an extremist voter makes their opinions dodgy, even if they vote for a mainstream party. However, if a BNP voter votes in effect for a mainstream party (as they are doing in having their second or third preference counted and their first, for the BNP, dismissed) then what is bad about that? They've voted for a mainstream party, good! Or are we supposed to screen all mainstream party voters to see whether they're not really covert BNP supporters who just decided to vote mainstream?
The argument that mainstream parties will court small extremist parties' votes only holds water, I think, in a system like that used for the Australian general elections where you have to list preferences - and even there, not that much water.
Why, in the UK, would parties risk putting off their mainstream voters by coming out with things designed to appeal to the BNP? Let's face it, even the Tories' current level of anti-immigration rhetoric hardly manages to charm the BNP voters, and if the Tories were to go significantly further right on this issue they would lose the more centrist swing voters - indeed, the sort of people who ARE likely to use preferences. So why spend that political capital courting the BNP voters - who AREN'T likely to use preferences? It would be not just a wasted effort, but counterproductive.