When you develop for console you develop for the lowest common denominator. This IS the advantage you have. Now you have two console with the same targeted environment, one slightly better at throughput than the other, and initially with lower numbers. Would you target that, add millions of dollars worth of development in bell and whistle , only to have barely a better sales, and risk the ire or their "lower" console brethen because the screenshot presented were only for scorpio ? And if you develop *for* scorpio you will miss on xbox one 1.0 player.
I know that some hold that console generation should be shorter or whatnot (e.g. totalbuiscuit on youtube argue that that the last generation was too long) but the splitting ehre will make it so probably nobody with a sane mind will invest into getting more power out of scorpio and having different config and development on an already expansive console development without real benefit sales wise (at least initially). And if nobody is adding bell and whistle, why should people buy the new one when the old one less expansive does as well ? Remember console fit one niche : the one where you get a hardware with known spec and develop for it. Scorpio break that. It will add confusion on buyer, reviewer, developper side.