Cube printout should not compute lazy scalar coordinates

[pp-mo]

To address #5723

We've accepted in the discussion there that we really don't want complex lazy calculations performed each time you print a cube.
Nor would we like it to "realise" a scalar coordinate, so it only happens once, since you really wouldn't expect printing to modify a cube.

It seems reasonable instead to just print something like "complex_scalarvalue_coord: <lazy>" in this case.
The user could of course realize it, if + when the value is actually wanted.

Note: IMHO this is not a major change, and can just be done
-- we have changed printout format before without considering it a breaking change.
Also relevant here : most scalar coords, ones loaded from files, are not lazy, since #5229
( so in many/most cases, adopting this won't change the output )

[bouweandela]

Does this issue mean that there is consensus that this is the right approach? And if yes, is this is something @fnattino and I could help with implementing and/or reviewing?

[pp-mo]

Does this issue mean that there is consensus that this is the right approach? And if yes, is this is something @fnattino and I could help with implementing and/or reviewing?

Yes, and Yes, I think!
I wrote this after last week's Peloton meeting, so we did get to generally discuss the options + this was our choice.

N.B. I just edited the comment as I realised the "<lazy>" part didn't come out in markdown!

[ESadek-MO]

@SciTools/peloton Closed in #5896