Do clouds reflect ambient noise? On days with a full blanket of low-hanging clouds overhead it seems I can hear more trains/traffic in the distance. Do clouds actually reflect ambient noise or is this some trick of the imagination?
 A: Clouds can form at a temperature minimum. Above the clouds the temperature may increase sharply.
The sound speed increases $T^{1/2}$, so immediately above a temperature inversion, there can be a region of decreasing (sound) refractive index. This can have the effect of bending sound waves back towards the Earth.
The phenomenon is more normally noticed on still, clear nights, where the ground layer can be colder than air higher up.
I rather suspect that this is a related question Why do I always hear remote train horn at night?
A: The propagation of sound over distance definitely changes with weather; but perhaps not the way you think.
For example, it is my experience that if you stand at the shore of a quiet lake (say 1 km across) at night, you can hear sounds from the opposite shore. This is a result of changes in the density of the air, which gives a "focusing" effect. Specifically, if the air right above the water is cooler, it will be denser - and this will slow down sound. The result is that the air column acts like a giant wave guide: sound that is aimed "slightly up" gets bent down again. And thus instead of the sound propagation being 3D (with inverse square law behavior), it will be quasi 2D, with $\frac1r$ behavior. Which makes the sound "carry".
Now if you have low-hanging clouds, I imagine there are also thermal layers in the air below these clouds. And it is conceivable that on some days that results in a similar constraining of the sound, and improved propagation.
Would a cloud reflect sound? The density of the air is greater, and thus there would be a acoustic impedance mismatch. But the surface of the cloud is rarely smooth enough to maintain phase relationships needed to keep the sound coherent. Perhaps some low frequencies would carry (where the wavelength is long compared to the irregularity of the surface of the cloud).
My money would be on thermal layers first, and clouds second.
My hunch is strengthened by an interesting article on "acoustic shadows" and their possible influence of the outcome of the US Civil War (see "Outdoor Sound Propagation in the U.S. Civil War" by Charles D Ross)

The battle, silent to Johnston two miles from the front, was heard clearly by citizens of Richmond ten miles to the west and to Federals as far to the east. The probable cause was a temperature inversion bending the sound back to the ground. On the night before the battle, a violent thunderstorm (many soldiers said it was the worst they had ever seen) raged over the area. The day of the battle dawned with widespread, low cloud cover-ideal conditions for a low- atmosphere temperature inversion.

A: The only way to answer your question is by means of experimentation.
As your post implies, the mind sure plays tricks on itself. And how we perceive the world is not how the world actually is.
So get yourself a decent sound meter and get measuring!
Should you find some effect the question then becomes "how and why?" That would be an interesting question for this site.
