When running the VL53L0 runs kind of hot. Unless it were really cold, I'd expect the sensor to warm the glass enough that it wouldn't fog up. But lots of people run the sensor occasionally - rather than full time.
I did an experiment where I used a particular type of coverglass, and covering that glass with a rather large amount of coffee creamer (to simulate dust) produced no false readings - but it did limit the distance at which the sensor could range.
Have a look at 閎喬光學 – Hornix Optical Technology
www.hornix.com.tw/en/home
Click on Product and search almost all the way down.

That coverglass on the right (IR079C0-IC09-A066-2) has a raised barrier of opaque plastic between the transmit glass and the receive glass. And it fits tightly over the sensor preventing any photons from short-circuiting the system.
With a glass such as this I'm hopeful the chip will warm the glass enough to keep you out of trouble.
With a glass dust and dirt might prevent you from getting an answer, but you will not get a wrong answer.
Note: Hornix is not sponsored by ST. But we did jointly design that coverglass.