Amplified is a better word. Dan doesn't think there are rocks in hills and shares photos of the cottage with a limited view of trees; there are better ones that show hills on all sides, perhaps I can post a better one. Perugia is surrounded by hills make up of calcareous breccias a type of rock.
This is an extraordinarily poor and unscientific "argument", without a shred of merit. You are plain wrong. 100% wrong. Perhaps a basic science lesson (I'm not sure whether it would be correct to suppose that your "argument" is the product of little or no scientific training or education?) might help.
Here's why your "argument" is so stupendously wrong: Sound is transmitted through air by pressure waves, which are in essence* formed by "squeezing" the air in a rolling wave motion. The energy transmitted in sound pressure waves through air is therefore entirely bound in the energy of individual molecules within air (chiefly nitrogen and oxygen), which gain energy as they are "squeezed", then transfer it as they "unsqueeze" and the next molecule along "squeezes" in turn.
With me so far? I hope so! Now, here comes the first important bit: it's the difference between sound and light. Have you ever perhaps wondered why you can see someone from 500m away, but you can't hear what they might be saying in a normal conversational voice at that distance? It's chiefly because light is high-energy electromagnetic radiation, and sound is low-energy pressure waves in air. The reflected light emitting from the skin and clothes of the person hardly degrades or fades at all over the 500m distance, whereas sound pressure waves degrade a huge amount over the same distance.
So, the first point to notice is that sound waves degrade and fade much, much, much sooner than light. That's important, but we'll park that for now.
On to the second important bit: What happens to sound when it hits a solid object (especially when compared with light)? In order for sound pressure waves to be "reflected" off a surface, it needs to hit a very smooth, flat surface in order for the air to "bounce" off the surface in an orderly waveform pattern. If the surface is not smooth, the "reflected" sound will get progressively distorted from the original sound pressure wave. In addition, the very act of hitting a solid surface degrades the power of a sound pressure wave substantially. This is a very different situation from light, where a highly-mirrored surface can reflect light radiation with virtually no loss, degradation or distortion. As a further factor, most sound pressure energy is lost if the wave hits the surface "head-on", as opposed to a more glancing contact (which would of course not reflect the wave, but merely deflect it by a small angle).
Furthermore, if the surface encountered by the sound pressure waves is a) rough-textured and b) not flat, a vast majority of the sound energy will be lost in the collision. That, for example, is why sound recording studios can use material similar to cardboard egg boxes to line the walls and ceiling in order to almost eliminate sound reflection.
So the second conclusion is that sound pressure waves cannot be reflected without a very significant loss of energy, even if they are reflecting off a smooth, flat surface. When one adds in trees and other foliage, the loss of energy from any reflection would be very large indeed.
In summary, sound cannot be "amplified" in a canyon. That is a specious and totally incorrect supposition. I am guessing that it may be predicated on a mistaken and incorrect assumption that sound pressure waves might behave in a similar way to electromagnetic light radiation. Whether such an erroneous comparison is responsible or not, the fact that remains is as follows: any sound pressure waves emitted within Meredith Kercher's room on the night of the murder - even if her window was open (and there is no evidence to suggest that it WAS open, since it was found closed) - would have hugely degraded in strength by the time they had radiated out to the valley below, reflected back up past the cottage again, and arrived at Sig.ra Capezzali's window.
In short, the elderly lady (and the other "ear-witnesses") were either a) honestly mistaken, or b) they heard something (a shout or scream perhaps) from directly on the road or the parking lot, or c) they are fantasists or liars. What they DIDN'T hear was a scream originating from within Meredith Kercher's room, let alone one loud and piercing enough to wake someone through double-glazed windows.
* The real mechanism is actually a bit more complicated than this, but my slightly simplified description is accurate insofar as it allows for a robust description of the properties of sound pressure waves in air.