this would likely be the best way to rapidly innovate solutions for layer 1
Layer 2 experiments are probably the only way to innovate solutions for Layer 1.
And, to be quite honest, I probably can't argue persuasively against the inertia that keeps Layer 1 from innovating. Getting something wrong on Layer 2 is only costly for those who dared to experiment. Making a dramatic change to Layer 1 and getting it wrong could end up very difficult (and perhaps impossible) to undue the unintended or unforeseen damage. With that being said, there are some changes (experiments) that have to occur on Layer 1, such as adjustments to the 'haircut rule'.
In any event, the key to implementing any substantive changes to Layer 1 (especially with respect to the social media aspects of Layer 1) will require 'proving' the soundness of those changes on Layer 2 first. Even then, there will still likely be hesitation to implement changes to Layer 1. The irony here is the fact that a hugely successful Layer 2 solution will probably mean the change is not needed on Layer 1.