Yes, the revisions clarify those points to me!
Additional suggestions to the revisions -
Comment A - Many Branches for a Tree
“invariably, at least sometimes” comes across to me as conflicting. Is it one or the other?
Maybe change that part of the section to read something like, “…deliberate attacks, it would not be uncommon or unexpected for the proposal mechanism to produce multiple children of the same parent. Casper’s job is to choose a single child from each parent. This selection decides upon and identifies a single canonical chain for the block tree.”
Comment B - Supermajority Link Definition
Adding the distinct supermajority link references to 1c clarifies my original confusion. I’d suggest adding r -> b3 as an example for completeness, if that can be a supermajority link.
You also may want to define “r” in the figures, rather than relying on the reader to make the link between it and the earlier definition of root or genesis bock.
I understand the point about skipping checkpoints. I’d suggest changing that section to read something like -
"Supermajority links can between adjacent checkpoints or skip checkpoints, i.e. it’s pefectly ok for h(b) > h(a) or h(b) > h(a) + 1.
Now, on to the rest of the document