since dist(x) = <n,x-p>, should it be ax+by+cz-d in this case? or we can have d to be <n,-p>

^I was also wondering that.

Also, isn't there still a correspondence between points and planes in normal 3D coordinates?