Okay. So what's wrong with that? It's like saying the difference between protected and unprotected speech can be the location of the speech in question. Yeah, it very much can be. What's wrong with that?

Because path of the file is not determined by the user in all cases. This means the prosecutor will have to prove that the user deliberately stored it for later consumption.

For example, if a web browser has a semi-permanent cache by default for all pdf files, and the browser downloads a pdf to this cache, it's not possible to determine if the user "downloaded" it (ie. deliberately, to access later or repeatedly) or "streamed" it (ie. accidentally accessed it and closed it after viewing the content)

The fundamental issue is that politicians and the general public do not understand how software works. Many applications access and store data by default which the user has not implicitly asked them to do. If the law is determined so broadly, it can put any user with an application that accesses content without somehow validating that it is legal (how you do that without accessing it I do not know) into a grey area which the state can choose to exploit/prosecute with if needed.

its not reliable? It can easily be forged/hidden/framed?

Again it confers power to those that understand how systems work. I'm not complaining, it benefits me but we continue to drift toward a class-based system where the technical class gains ever increasing benefits over the rest of the plebs. We're sleepwalking into an increasingly stratified society.