I agree with most of what you said with the exception of your idea of 90% of portal players couldn’t hack it at their previous school. I would roughly estimate that number to be about 40-50%. The real story is so specific to the situation (poor coaching, attitude/work ethic issues, scheme fit, etc).Good subject matter. My theory on roster build, is that you don't want to go to the portal every year to find players to fill your roster. From a HS recruiting aspect you should keep your class sizes fairly consistent and inline with distributed position choices. Invariably you will have years where there are a large numbers of better athletes at certain positions, when that happens you must adjust how you divide classes going forward, so as not to be out of whack in your distribution.
You should only be going to the portal when you have fundamental failure in your recruiting of positions or unexpected injuries or portal emigration (players leaving).
General rule on the portal is that 90% of portal players are ones that didn't prove to be talented enough at their original school. Another 5% are complete mercenaries, who won't be devoted to team goals. Only 5% are probably worth your time and effort, and that is going to be expensive and require a lot of effort to cull into your roster.
It is more productive to develop your own players and keep them.
10 to 15 years ago this process was pretty well understood by both schools and coaches; then the pay to play decision came down, and the portal happened, and everyone, and every school simply lost their minds on how to handle their rosters.