I think the idea is that he is looking for horses who have shown either the Red-Scan moves or interior moves to be the contenders because it "energizes" them.
He is not eliminating horse, he is including certain horses. He got the field of 10 down to 7 and then further refines it. On page 13, he outlines his whole procedure.
The first two races in the book were stripped down examples of the technique. In both races, I see that a horse who won last out with a top Beyer failed to hit the board at relatively low odds, so his idea of looking specific horse worked for him. In his newsletters, he covers other ideas and how he uses them, such as speed,fade. When he sees the other moves, he focuses on them and not the red-scan like he did here.
At CJ's old board, we had a guy who looked for all the red-scan horses every day and posted a list of them. I take his ideas and incorporate them into my own play, like I do with those of Randy Giles.
As far as the two horses he ignored, I think in maiden claimers, losing 4 in a row while losing ground in the stretch every time, and being the top two betting choices is a great idea. Would you seriously bet either of them at the odds they went off at? I am not sure I would ever bet the winner, unless his final fractions looked strong, but at least the race was opened up for a good price.
And, btw.....the winner was trained by Jason Servis.....
Oh, and the red-scan technique pointed out Accelerator in the Pacific Classic today.