"The tools to generate pin configuration and clock configuration are ideally only used at the beginning of the definition of your project. Once you have all the pins and clocks properly defined and configured, you can focus on develop your code using the KSDK 2.0 peripheral drivers or middleware stacks (USB, Filesystem, etc)."
That never happens in real life. I've had to change up peripherals and configurations of things late into the dev cycle because of finding limitations and problems that I've had to design around. Assuming that you know exactly how the hardware is going to work prior to doing code is not what reality looks like.
To be fair, the software industry had the same unrealistic expectations 30 years ago. There were application generators that took specs and then generated stuff from it assuming that the creation is the problem, not the maintenance. It's always the late stage and maintenance that kills you, not the creation. The extreme example of this is 50 year old programming: flow charts. It took about the first line of code to make those irrelevant to what was actually going on.
Don't get me wrong -- in the hardware realm, PE (and the tools to come) get you up the learning curve of the part a LOT faster so are invaluable to those of us who don't spend our time being an expert on this one part (I've got several manufacturers parts to know well enough to use!) and so are useful even in the start of the project. That said, integrated ones are WAY more valuable than starter ones.
So, if there are NXP tools team members listening, let them with ears hear - make those tools part and parcel of the larger tool. Cypress PSoC Creator is the Gold Standard for integrated tools (with the exception of not knowing about RTOSs). There is no tool I've seen better than that one for management of parts. Emulate that. Splitting up PE is a step backwards in my seasoned opinion.