I like the feature request, but is it? How many people are saying "I think I'll switch to Tidal" over a subtle feature like this? How many people prefer the faked shuffle way?
What are the odds that Spotify originally did a true shuffle, then at some point A/B testing "proved" that the modified shuffle improved some metric? Not to mention that putting a thumb on the shuffle function lets them manipulate plays and maintain the sweet payola revenue.