Why did the WPA analysis of the Adelaide Test not reflect how tight the match was on the last day?
Good point. See our explanation here.
In the Adelaide Test, Graeme Swann took five wickets in Australia's second innings, so why wasn't he credited with a higher contribution to England's win?
He did indeed take five wickets, but he also bowled more overs than anybody else in the second innings. He was therefore more responsible than any other bowler for allowing the Australian batsmen to stick around at the crease, thereby making a draw more likely. See further discussion here.
Why doesn't the relevant fielder get credit for the run outs?
We decided to keep run out numbers separate from batting and bowling numbers. We do this because it is not always easy to attribute a run out to a single fielder. For example, the run out of Xavier Doherty at Adelaide contributed 3.5% to England's chances of winning. But to which fielder should that be credited? Strauss, Cook or Prior?
What is the formula that underpins the calculations?
It's not a secret, but it will take a while to explain. Keep checking back at this site, and very soon we will put up a complete explanation of how we calculate these numbers. See here for a brief introduction to WPA.