I'm surprised there is no universal and definitive opinion on what is essentially a simple accessory. While possibly not offering much protection to the wearer I guess it makes sense to wear one while using public transport if it in any way helps limit the spread. I'll trust the Germans on this one and would have no objections to seeing the same here whenever some restrictions are eased.
There is - surgical (and anti-pollution type, or home-made barrier type) masks work to a degree. A low degree.
They work best when an infected person wears one, and coughs or sneezes (or breathes but that's a much lower risk) in a public space, in which case the mask will significantly shrink the cloud of infected droplets. This is its job.
People knowingly infected / symptomatic shouldn't be out in public spaces. We don't have enough data to suggest how many people are shedding virus before developing symptoms - but it seems to be much less than originally feared, and for a maximum of a day or so.
In terms of protecting the wearer from other people... they're really a bit sh
it. Is an infected person sneezes right in your (masked) face, then
IF correctly worn and correctly removed, it should reduce your viral load by about 50% (as opposed to 80% for a standard dust mask for diy or 95% for PPF2 level recommended for any medical person working within 2m of an asymptomatic patient).
Quite honestly, any barrier mask will work as well as a surgical mask, so long as it fits snuggly. The price gougers have driven the price up by 1500-3000% (£10-15 used to get you a box of 250, now you see a box of 50 going for £70-150). You can still get cycling / anti-pollution type masks for about £15 or activated charcoal (no better than the previous for this) for about £25, which will be about as effective as a surgical mask, whilst being reusable.
You don't seem to be able to get FFP rated masks (even FFP1 rated) for love nor money, even as a clinician.
If that was the whole story then (stocks and price gougers allowing) everyone should wear a mask of some kind. But. The problem then comes with the false sense of security.
Everything we know about human behaviour shows us that if you give someone the illusion of security, they increase risk taking activities, be it seat belts, scrum caps or face masks.
In this case, it means that people wear a mask, which they think is effective, and A] don't wear it properly, rendering it way, way less effective B] touch their face more, leading to greater spread from their hands (even if gloved*) C] actively keep pulling at the mask, rendering it almost completely ineffective, and D] go out when they shouldn't, stand closer to people than they should, spend longer in places than they should, etc etc etc.
*Gloves are also pretty useless the way most people are using them - they are most effective as a visual cue that you're about to touch your face, and so deciding not to.