Now obviously these masks are not the most effective masks in the world. However, is there any cog sci evidence that they provide any protection at all? I highly doubt there's any evidence to show they prevent someone from identifying you, but is there any evidence to suggest they even slow identification down or increase errors?
I don't know of any proper identity recognition experiments using such a mask, and obviously it depends on how difficult the recognition test is. However, here is a paper where they found that sunglasses (and facemasks) impeded accuracy in an emotion recognition task. I would guess that often identity recognition is just as difficult as emotion recognition, so I would assume glasses or a mask would make people (slightly) worse. In general, this is a question about what features are necessary/useful for face recognition. There is lots of research about this, and studies such as this show that the eyes are particularly important. So disrupting this part of the face, and potentially the relationships between them, is actually not a bad way to disguise yourself (though obviously a full-on balaclava would be better!).
However, is there any cog sci evidence that they provide any protection at all? I highly doubt there's any evidence to show they prevent someone from identifying you, but is there any evidence to suggest they even slow identification down or increase errors?
No and yes. As far as I know there's no evidence that masks in particular provide a special protection that differs from other physical objects that disrupt the face/bodyparts like when you cover your face with your own hands.
However I think it is helpful to explain (concisely) how identification and visual recognition in the process of perceptual organization works:
Stage 1 When we see something in our environment (face of a person) our organ of perception (eyes in this case) receives a stimulus (light) that causes the creation of an neural impulse that is send to the brain (image of a face).
Stage 2 Now our brain combines the information (the image of the face that we just saw) with a-priori-knowledge from our brain (a cocktail of experiences, expectations, language knowledge...). By doing so it identifies the object (face of my friend) and recognizes it. Recognizing means you are aware of the object's function (it's a friend, not a foe -> no danger).
Stage 3 The result is a mental re-creation of the object we just saw – a so called percept.
So to sum it up, if a mask or any other physical object is in the line of sight, it alters the information we receive about the covered object. And if our brain cannot fill in the blanks, the stage of identification is slowed down or the object cannot be identified at all. So it depends on what part of the face is covered and to what extend and how much information about the object we try to identify is already available in our brains.