Speedier Import: identify duplicates by filenames/timestamps (without reading the whole RAWs) Merged

Situation: With ≥128GB cards, it's tempting to leave older photos "as a quasi backup" on the card, just in case... With <10 to 30 GB per shooting this card could hold 7 Shootings until you need to clean up (or rather change to the second card).

So when opening the import dialog:

[1] you have to wait for capture one go through the whole list of photos on the card, even when 'ignore duplicates' is checked -> they repeatedly appear for a split second and then disappear which can take quite some time for several hundreds of photos...

After benchmarking my cardreader it seems to me CO reads the whole RAW-data for identifying duplicates and after that has to read the whole RAW again when performing the import??

Suggestion: maybe just read just the filenames/timestamps and not the whole RAW?

[2] quick selecting "today's" or yesterday's photos by simply ticking a single 'date'-checkbox could speed up this process.
I just realized: there is the Filter Tool in the importer/cull sidebar. It sort-of does this job, but it does not select the pictures so you have to "select-all+spacebar+enter" to import the filtered list.