In addition to cattle, I (rarely) breed dogs. For our last two litters, we had puppy requests from the US, Canada and several foreign countries (UK, Germany, Belgium, South Africa, Argentina, New Zealand and Australia). I'd like to think we have good dogs and we make sensible decisions when breeding.
Some of the people in that endeavor who have reputations I value have routinely practice the habit of tightly breeding for two generations ... then an outcross ... then back to your original stuff for two more generations ... then out again ... then back in for two ... etc ... They stress the importance of maintaining a keen eye for problems and issues, without sacrificing phenotype, and be uber selective when it comes time to decide those who get to perpetuate the gene pool.
*******************
On to the topic of flushing females for embryos ... I agree with you (generally) that you don't want to flush a cow unless she's into her teens. There are exceptions, but they're ... well ... exceptions and not the rule. I don't know that I understand why go through the cost and expense of a flush if you're only wanting terminal traits; unless I am misunderstanding you, I'm not sure how that pencils. To me, it seems to make the most sense to use the best "cow maker" bull(s) and make as many cows as you can ... a series of full and half sisters.
The intent of laying out the original scenario was to get others to give their thoughts on "if you were going to build a seedstock operation from scratch" ... basically, if you had a clean slate, what would you do in terms of selecting your initial animals and growing their numbers from there. I intentionally didn't say that the objective was to have 200 head of mama cows and that, to start, you go out and buy 200 head of mama cows from someone you knew and respected.
Presume someone wants to start somewhere and grow, what do they do?
...well, they go somewhere, and buy some cattle they like and they start...
Personally, if I'm starting a seedstock operation from scratch, I'd rather have 5-10 awesome cows (i.e., proven animals from proven herds) that come from systems and environments that are as close to mine as I can find and breed them to the best bulls I can find (AI or otherwise) who are also developed in an environment as close to mine as possible ... than to ... try to build an elite herd by purchasing 40-50 question marks at the sale barn and then spending a decade trying to get them where I want them.
And, when I say "best bulls", I mean that to be a general statement ... most of us here aren't dummies and yet we have different opinions on what "best bull" means. Best bull for you and what you're trying to do.