Great work, Yobbo.
Anyone who looks at that pic and calls it a fair delivery is deluded.
One of the great travesties of the game is that this chucker is still playing, and every umpire who had the balls to call him has been sacked.
“The anthropometry assessment clearly shows that Mr. Muralitharan has a natural 35 degrees of elbow flexion during standing, which during the delivery action (under load) reduces to a value of approximately 24 degrees”
So, we are back to where we started, more or less, taking into account the internet’s ability to garble things, and the fact that the measurements probably have 3 degrees or more of error.
Well, I admit it is strange to me to measure an angle in percent, but it is reasonably clear why it should be a percentage of 180 – 180 degrees is flat, no bend. The most possible bend you could have (with an infinitely thin arm) is for the forearm to be bent all the way back, parallel with the upper arm – a bend of 180 degrees, or 100%. Darryl Foster, the source of the 21% quote, says 38 degrees in this interview.
Now, Foster was one of the authors of the study you linked to, which gives the angle as 35 degrees, so I don’t know what’s going on there.
As for the reduction to 24 degrees during delivery, that’s weird. The bit you quoted suggests that every time he bowls, the elbow gets as straight as 24 degrees. I think the authors were lazy when they wrote that, since their data explicitly shows that the suggestion is false. If you look at Figures 2 and 3, the angles it gives all over over 30 in Figure 2, as high as about 35 (you need to subtract the values on the y-axis by 90, otherwise it doesn’t make sense), and between 30 and 24 after the remedial work. Only in one ball out of the twelve (or maybe as many as three, the picture’s low-quality) did the angle get as small as 24 degrees.