Re: Would something like an temporary protective bubble help?
Let us assume your shell can be deployed in 1 second (using no more than solar power and not completely destroying the alignment of the telescope due to vibration, so this is pure fantasy). These objects are travelling at upwards of 10km/s and perhaps are 1mm across. So you must detect it 10km out. You will need a telescope to do this, which we assume will use visible light, say 500nm wavelength.
So Rayleigh criterion: we need to be able to resolve this thing. theta = (1/1000m)/10000m = 1E-7. lambda = 500E-9m. D = 1.22 lambda / theta which is approx 5m. So to see this thing you will need a telescope with a 5m mirror working in visible light.
JWST main mirror is 6.5m so this telescope is smaller ... a bit. Oh but you must cover the whole sky with this: you will need many of them pointing in all directions. Let us say 6?
But of course you will need multiple observations to estimate trajectory. And spotting it 10km out is hopelessly later than you must really spot it, perhaps you really need 100-1000km away. So say 100km now D = 50m. Perhaps you could work in UV make it a bit less? Say 25m dishes, 6 of them.
Think of the kind of missile detection and tracking systems warships have, but the objects you wish to track are travelling at mach 30-100 or more, are the size of rice grains or smaller and have no convenient hot exhaust. And they can come from any direction in the full sphere. And the system must run from solar panels, must run cold to not disturb the telescope, must emit no vibration and have no real moving parts (momentum, angular momentum) and can not emit any kind of exhaust or anything which might land on the mirror or optics.
Yes this is not even slightly possible