It seems it can be great idea to make new curve based on open and peer reviewed standards.
There is no need to do that. There is nothing wrong with or even slightly hinky about the curve we are using as discussed in detail in the thread described in this post: