@User The restrictions are come from a couple of angles -- the two big ones are streamlining the process (as @User pointed out, we have a very small team working on this) and optimizing for metrics different from typical industry approach (aiming to provide ICs back "ready to go", getting a large number of ICs back so your prototype isn't a precious "one of a kind" thing you can't share with others, having a very wide range of different user types, etc) -- I go through some of it in the PDK introduction talk

All fair enough, especially where it simplifies the work of building the reticle. And I appreciate the kind offer from @Tim Edwards to at least consider swapping out some pads where absolutely necessary to achieve performance. I guess I was just trying to find a way for users to do this without him having to become involved every time this is necessary. Anyway I need to spend more time on designing rather than complaining, otherwise I will not get to the stage where this is relevant.

@Chris Jones: It's one of those things where I will probably have to do it once to determine what's the best way to automate the process for the next times it's needed.