Isn't this just increasing the branching factor? How hard will it be to find new melodies a million years from now? Easier than today, the same, or harder?
It is, but it is increasing the branching factor by much more than finding all the "good" melodies in 400 years.
Here's a good example of music that stands on it own right despite using the melody of despacito: https://www.youtube.com/watch?v=ydqReeTV_vk - I'm pretty sure that most people at HN will appreciate this one whether they like despacito or not :)
I think the relevant thing about the pigeon hole principle is that it deals with fundamentally discrete things. Sure, there is a finite number of scores using twelve notes in a finite period of time, but that's just not how we judge or compare music. I would say that there's so many branching factors in music that quantifying one piece is akin to measuring a coastline; the smaller your scale, the longer the coastline. Beyond timbre, there's other cultural factors like, who wrote the music. In a million years we might have umpteen songs with the same score, but they'll still occupy different pigeon holes in the collective consciousness.