As others have said, import agent is great for bulk import, especially if the tif files you are importing are large files. 10,000 pages is not too bad to churn through if they are all 1 page docs but 500,000 pages is another level if they average out to 50 pages each.  

Personally I like to combine that in situations like this with Distributed Computer Cluster. You can have import agent pull everything in and add a tag such as "OCR Needed". Then you can have a workflow that periodically runs adding documents to the DCC Queue by doing a search for the tag and only adding the first X of them.

This way if it is a very large amount of documents you can add more machines into your cluster to churn through them faster. Plus the documents are available as soon as they've been uploaded, even if they haven't been OCR'd yet. You just have to watch out that you don't have too much in the queue at once. It runs into issues if it has more than 4-5k in there at once. (hence the workflow that runs periodically to feed documents in at a slower rate).