I agree that it represents progress, but I believe it should have been implemented as a hardfork to fix the very "softfork" security issues it takes advantage of and without the "lighter" "full nodes". However, my point above was that another of your anti-hardfork arguments isn't holding water.
That is a misconception, there is a very good reason why it is not implemented as hardfork.
Besides this is the only smooth way to transition, otherwise we would risk the entire network with a synchronized hardfork, that is too dangerous.