IRIS experiment is not useful

[NovemberOrWhatever]

The IRIS experiment has a massive data size and a tiny science return, and is not exactly good in terms of power consumption, cost, weight, or duration, either. Therefor it makes little sense to ever use. I'd suggest multiplying the science return by 3–5 or decreasing the data size to make it make sense.

[Johnkurveen]

That certainly makes sense, though I actually like having an experiment that's... nearly pointless. It's a stretch goal, a "wow, I can even include IRIS". And it teaches us the value of actually setting a requirements document and doing the trades we should be considering, rather than just saying "I want every experiment"

A nearly useless experiment provides interesting choices, so it's better than just another decent experiment... in a way.

(Someone who has put hours of thought into this exact issue)