This is not in our short or medium plan. We understand the need but don’t feel like it’s our priority right now as we focus on the adoption of the modern workplace.Dave Kuehling commented
For me it would be based on more than just the metadata, but an actual bit level check. I have found, especially when looking to pull in content from file shares, the file content will be the same, but the name, date stamps, and other things could be different, and will likely be in a different folder. I think the data points you mention are good to use as well, but having that bit level check will truly identify dups.