Your math assumes you are the only person using masks and have daily exposure.
It breaks down when people are in isolation, which many of us now are. You are not out daily, so a 50% reduction is a 50% reduction, and with luck you never accumulate enough to get sick.
If everyone wore masks, then every ones daily exposure would drop dramatically. You get 50% less intake, and infected people are emitting much less too. You have pretty much eliminated air born transmission, and can now focus on hands and eyes.
There's a psychological component to this. If a mask reduces the chance of infection (from daily contact) from 60% (random number) to 54%, I'm not going to bother, even though in aggregate it might make a huge difference. That's just me though.
Also, I suspect half-measures like this are only effective in the early stages of the outbreak. For instance, if the viral load required for infection is 10 units (completely arbitrary), and your daily exposure is between 5 and 15, using a home-made mask that filters out 50% of viruses makes sense. However, if your daily exposure is 100 (presumably because everyone around you is infected), that 50% reduction isn't going to help you.