There are good studies on viral load reduction from masking with flu viruses from before 2020. After 2020 the topic became politicized so the studies are probably unreliable.

The assertion is that if you dump a gooey grey gallon of pure covid on someone, their symptoms will be more severe than if they only get 18 virions. Do you suppose that wouldn't be the case?

From that link:

Our simulation also shows that wearing a mask can effectively reduce the spread of the viruses.

Sure, by some metric masks are very effective at stopping viral particles. After all, in my example the mask will stop tens of thousands of viral particles from reaching you every minute.

Another metric is, will that stop you from getting sick? The threshhold is only 18 particles!

All the factors you mention are scalar. You can wear 0,1,2,3 masks. You can be 1,2,3,10,20 feet away from the person for 0,1,5,10 minutes. I can come up with any number of such factors involving wind, lifetime of virus, the person's shedding rate, etc.

Given that, there will always be a border at which having an additional 50% coverage is useful. Where the actual line is, we don't know. But as long as protection is monotonic in the factors above, it's always valuable to have more protection, which implies even <100% effective masks are useful.

Trivial proof: Imagine someone with literally 10k masks all around them. You have to admit this is more safe than 1 mask. QED. You can call in meta factors like "in reality nothing matters" but you're arguing by consequences then, not actually disagreeing with the facts, just claiming they don't matter.