Using new AI bulls, and different ones and especially the unproven, new, great one, is like a hunter who likes to hunt many species of game. There is the thrill of the hunt, the seeking of the record book trophy, a need for new, a rush when things seem magical. Where AI is most useful, for the long run or a long term breeding program is in the proven sires that you know work in your environment regardless of how you sell or how different you want your cattle to be. But that is dull. The older bulls have proven EPDs and experienced opinions of offspring while the new ones rely on the genomic testing to try to get them close to the levels of the older bulls.
So, it is a game, no matter how you play: alike, different, outcross, linebreed. But the thing that is best is to win and the winning is based on the odds of knowing what the next card will be in cards or the tendency of the game. The new AI bulls are teen girls in mini skirts to an old married man. The low odds are the reason that Angus have plenty of bad feet, bad udders, bad disposition, low fertility, .. and weaning weights have not gone up appreciably on the national average in decades. Doesn't sound like a winning playbook to me unless it is all for a marketing reason away from the salebarn and meat industry. But, it is just money!