What is a bit unclear which isn't addressed in the original blog post is how they actually were able to successfully scan over all the data and copy it to the n

What is a bit unclear which isn't addressed in the original blog post is how they actually were able to successfully scan over all the data and copy it to the new cluster, in spite of corrupt sstables. CDC would not handle back filling historical data. Would expect that to fail for certain token ranges. Perhaps they ended up deciding to discard them?