It depends on quite a lot of stuff like what platforms specifically you're talking about, so there's no universal answer to this stuff.
The more general answer is that each added memory channel adds more memory bandwidth capabilities since each channel can access that particular section of memory while another channel is accessing memory. This means that if you have a single stick that can have ~30GB/s of bandwidth, a dual channel setup will have ~60GB/s and a quad channel setup will have ~120GB/s (in theory at least, in practice there's a lot more factors involved with this). Some workloads love memory bandwidth and therefore love quad channel setups, while others couldn't care less. Plus because of some weirdness in the way the memory controllers need to be designed, quad channel mode also tends to introduce memory latency that some workloads would not like (these aren't common, but they do exist).
Granted, some CPU architectures are also more or less memory bandwidth limited than others, and with the advent of DDR5 there's so much memory bandwidth that going to "quad channel" (technically consumer platforms support quad channel since each memory stick is two channels, but everyone just refers to each channel pair as a channel) won't have a ton of benefit compared to DDR4 and earlier setups outside of specific use cases.
Memory channel count shouldn't really be that high on your list of things to look for in a platform, since there's usually other factors that impact performance more than the memory config. While more do help to a varying amount depending on a lot of factors, they aren't the be-all end-all for determining performance.