Hi John Wilcox
I quote (emphasis mine):
Your targeted phase ( if you self manage ) takes as long and as many devices as you need to make your full deployment decision.
With 1803 is was 250 million and is documented https://blogs.windows.com/windowsexperience/2018/06/14/ai-powers-windows-10-april-2018-update-rollout/#FqIAJr2q01PQp6eQ.97
This continues to highlight the problem with your narrative. Microsoft releases SAC-T to 250 million devices, but it's not a release, it just some preliminary milestone. At the same time, Home edition owners don't self manage updates. Microsoft manages updates for them, hence 'you' is not [completely] applicable here.
That's why I'm saying that SAC-T was just another testing ring. Certainly, it's not in your narrative and perhaps, it's not even on your mind (albeit it should be). Yet this is the reality. The known issues sections for SAC-T as well as update history pages for subsequent CUs used to tell and continue telling the story of multiple issues being raised immediately after SAC-T (e.g. 1809). Thanks to all these testers who you pretend self manage updates :)
I'm sure you know how many of those 250 mil devices getting 1803 were running Home editions. Would you like to disclose this information?