Key to his proposal is the real life example of "Moneyball" - where a team proved it could use cold-hard-metrics to dramatically improve upon the old gut-instinct-only system Steve thinks this can be done for startups, and I am inclined to agree with him.,
I invite you to read the full text on his website and engage in a dialog about it. Could this give angels a much better way to pick the winners out of the super-early-stage companies we love to work with? Might it be a better fit for angel groups outside of the major innovation hubs - angels that don't already believe they "got it" :).
And lastly, might it imply a different model for angel groups? Instead of bringing in a few companies a month for "1st dates" that emphasize pitching skills, should we be bringing in batches of companies and putting them through a Lean Launchpad experience and evaluating them after 10 weeks of hard work (that we witness)?