What you describe should be termed “substandard” with regards to testing standards, not “fake”. As an example, fake would be further labeling the mask with a 3M logo.

I’m not sure how far you can take that. If they’re so substandard that they might as well be PM2.5 masks, are they still substandard KN95 masks?

Under that assumption everything could be substandard and precious few things would be fake.

From my perspective as a person who's career has been in manufacturing, from the factory floor to supply chain consultant to factory owner who lived in China for 10-years, there is a distinct difference between definitions here of substandard quality and fake. If some company is trying to sell a non-3M designed and manufactured mask as a real 3M mask, that's the definition of "fake". If a company has poor manufacturing practices but is using materials which would otherwise be acceptable industry materials to make a mask to KN95 standards, but for some reason, a random sample of the lot is proven to be "Substandard" to the KN95 standards, that is the definition of "substandard".

Yes, "fake" implies lying and fraud, and not just manufacturing defects. It's an important distinction.

A mask labeled as N95 that knowingly doesn't stop 95% of particles is a fake N95 mask. It is a real mask, it's a fake N95.

There's a big distinction between "we had a manufacturing error" sort of scenarios and folks who knowingly produce stuff that doesn't meet the standards (or don't bother to check).

If the consumer isn't lied to about the product, it's not a fake, even if the quality is bad.