Actually, I love this question, even though it has been debated several times. Basically, it comes down to your risk tolerance. Most people don't filter. I do, as do some others here.
For one thing, most people don't test for sterility:

That is actually considered an acceptable amount of endotoxin by some. But filtering helps remove stuff like that.
Many will argue that subq injections will face the body's own defenses, which is true. But, as
@hexagonal once pointed out, even subclinical infections can increase the risk of arthritis down the line:
You are a fat clown, here to attention seek rather to provide information. Here's a study showing that about 25% of vials get contaminated on repeated access: https://journals.lww.com/jehp/fulltext/2018/07000/A_pilot_study_for_evaluation_of_knowledge_and.120.aspx Now you are calling people...
glp1forum.com
Also, the recon process itself can create particulates that would ideally be filtered out. And when things are tested for purity by Jano, they filter first, so that their machine doesn't get messed up.
A primary argument against filtering is that the filtering process itself can introduce bacteria from one's home. But wearing gloves and wiping things down with alcohol can go a long way. So I don't really buy that argument, especially since, emotionally, international bacterial from Shanghai seems scarier than bacteria from my home.