As is, it seems to suggest that, after changing the data, a difficult hash for block A cannot be calculated. This is not true, right? I wonder if you meant “as the data associated with block A changed, block A will be invalid as the previous hash for block A may be invalid or not difficult.” and, hence, block A will have to be mined again.