The entire problem seems to be that the College Board (and ACT and SAT) took an initial design that works fine in a proctored classroom and applied it to a global testing scenario where it doesn't fit.
> If I were to design these assessments, I'd have it be an untimed, synthesis-based assessment...
Yes, but that takes more resources and has more variability among reviewers. It does have the advantages of being more fair for students with time zone issues or closures, and of being durable in the face of temporary website issues. Judging by their decisions, the priorities of the testing agencies seem to have been taking out the cost of the human element of reviewers (multiple choice is easy to grade, synthesis essays need expert reviewers) and the variability of the tests and reviewers to certify their results for schools.
> If I were to design these assessments, I'd have it be an untimed, synthesis-based assessment...
Yes, but that takes more resources and has more variability among reviewers. It does have the advantages of being more fair for students with time zone issues or closures, and of being durable in the face of temporary website issues. Judging by their decisions, the priorities of the testing agencies seem to have been taking out the cost of the human element of reviewers (multiple choice is easy to grade, synthesis essays need expert reviewers) and the variability of the tests and reviewers to certify their results for schools.