Using Technology to Avoid Stealing Songs By Mistake

In 1960, The Chiffons published He's So Fine:



In 1970, George Harrison published the song My Sweet Lord:



In 1976, a lawsuit ended with Harrison having to pay over $500K in damages to the original song's publisher for subconsciously plagiarizing (ultimately, Harrison got the last laugh by ultimately buying the publisher).

Just recently, Joe Satriani has taken Coldplay to court for the same infringement for allegedly copying his original song If I could Fly.



So my question is, could there not be a technology solution to this problem? With online radio sites having matching algorithms based on wave analysis, chord progressions, etc. could there not be similar analysis done to create a checksum hash for every published song that if matched against a new song would be deemed too similar?

It would be a lot easier than relying on someone's subconscious, just purely random coincidence or some judge's interpretation of what music is to award royalties.

Or even better yet, we would go back to the days of Mozart, Bach and Vivaldi when musicians copied each other liberally and were able to advance their art without fear of reprisals from copyright lawyers.