So I would hardly consider this a priority from an “APP” perspective or a “problem”. You don’t think it is a problem because it is only a problem for PI users. With 90% of the future users being PI users it would only be a problem for 90% of APP users.
Kees, indeed, I couldn’t state it more cleary, the problem starts in PI and luckily it is very easily resolved by either
1) saving all files in fits in PI (which isn’t harmfull in any way ? to your data )
2) convert your existing xisf to fits, which is also very easy in PI.