I understand that 16GiB (16*2e30 bytes) chip is present (delivered to user), but some of it is broken and not available (industry standard). Therefore capacity is less than 16GiB. Also I believe that there is a threshold, that manufacturer is using to throw away chip that are too broken to be sold. Ok so far.
So proper declared capacity should be threshold value (minimum) or threshold to maximum value declared range. I understand that this proper capacity is not very nice round number.
Problem has past: Around 2000, manufacturers started using GB (10e9) instead of GiB (2e30) that gives them some margin (7%) for chip defects. Later margin was not enough, so they started using more misleading specs like this. It seems that margin is at least 10% nowadays instead of <7% in the past.
Personally, I would be satisfied with stated margin size (like 7%, 10%, 12% etc) to maximum capacity, f.e. "USB flash drive 16GB max capacity, at least 90% usable."