For protection, besides visual confirmation of existence of protecion circuitry, one can test functionality of over/undervoltage and overcurrent protection.
As for fakes, some are so good that you may need nontrivial testing to notice the difference in performance, e.g. accelerated lifetime testing to notice the quicker rise in IR that we saw in the above graphs. But some important things cannot be (nondestructively) tested, e.g. construction quality. For example, manufacturers of fake cells typically have much lower quality control. Many ventings are due to metal contaminant particles which eventually lead to internal shorts (and due to cost constraints, the fake versions may lack features such as ceramic separators that help to safely control internal shorts).
Not to mention that some rewrappers use low-grade versions of top-tier cells - which would normally be rejected by most reputable vendors. These could have performance and/or safety problems (e.g. they may have been from batches that were recalled after contaminants were suspected to be the source of ventings, such as in many of the well-known laptop battery recalls).
