Updated the solution.

Man, I just destroyed two hours of work on another kata in a glimpse. It all started when I forgot my mouse at work. When I was working on the kata, I had to use my laptops' touchpad, which I am not used to at all. Then suddenly, I must have done a finger gesture which made my browser turn back the page - and all was lost. Now I cry like a baby....!

Thank you for reporting the problem! I think I fixed all issues. Sorry for the inconvenience!

I took your solution as complete working solution.

I have a Factor translation pending which tries out a new method of testing. I do two tests:

1. does the users result match a reference solution's result when evaluated for a handful of values of x
2. are there any further simplifications that the user has missed

And if both are passed then the test case is considered passed. This can become somewhat finicky as certain expressions can result in all kinds of invalid, or rediculously large values, but with care I think that it could be the best way to test this kind of kata.

Would it be unreasonable to just expect the unchanged order of the input for this case? Should the kata be changed ( some simplifications are already explicitly not expected ), or should the description just be expanded?

Is this a language-specific suggestion, or do you want this to apply to all languages?