I don't understand, and perhaps disagree, with the notion that a subsequent step in a transaction is associated with removal of the prior step.  That seems tidy, but problematic.  It seems more reliable to start from the assumption that each step supplements the prior state.  This is also how many systems (including legal) currently work.  As in our example of the life cycle of a (paper) check:

Of course, it will sometimes be useful to collapse prior steps into a subsequent one, not least when the prior steps are drafts or erroneous.  And the right to be forgotten or efficiency will require purging.  There are parallels to this in git, which assumes retention of prior states but allows collapsing and deletion. 

That quibble aside, the explanation seems very clear and focuses on the essential advantage of "DLTs": common record formats and granular synchronization.

--