Deduplication would become a whole lot simpler if I could match transactions on the actual time they took place instead of on the date. My banking data has the time in the meta data, but only the date is imported. Due to an issue in reconnecting to my bank, I have duplicates of all transactions imported over a period of 1 year. A simple deduplication based on transaction time would have been 10 seconds work, but now due to date and amount being the only matching criterea available, it now also highlights all other transactions with the same amount on the same day. For instance, I get the same cup of coffee 3 times a day at work, so these are all flagged as duplicates, whilst they are not. I now manually have to review every transaction, which is a pain. I could also imagine that this would be resolved in a more extensive deduplication feature where I can also filter / use invisible data (like all the data fields available in the "Imported Transaction Metadata" imported from Plaid.