In the process of analyzing the data I focused on inserts/deletes as that is how our system are designed, in the actual migration on the critical high volume DBs I synched all data using full backup/restore + Tlogs. These critical DBs are where the business makes money so couldn't lose any data. On the less critical or lower volume tables I used a one time backup/restore weeks in advance of the migration followed by redgate data compare scripts - DC allows you to customize what data to synchronize. These scripts are the key as they allow you to setup jobs to synch hundreds of tables. No data was lost other than several tlogs failed to restore at 2 am.:-)
Note that tables without primary keys or unique indexes need a bit of testing in Redgate to ensure they are coded correctly.
I spent a lot of time analyzing our data insert/update/delete/select pattern before deciding updates weren't a significant part of our production process. Your site might be very different.
Recommend you plan for missing data and let management know what to expect - it took about a week or so to clean up security, indexes and some less critical reporting tables - but the business was up in and making money in just a few hours and that is what I was after.