[Issue]: Dataset transforms unclear about being a list or object

[btbest]

Describe the issue

Each object in datasets MUST contain the field coordinateTransformations, whose value is an array of objects that define a transformation that maps Zarr array coordinates for this resolution level to the “intrinsic” coordinate system. The transformation is defined according to transformations metadata.

And further down:

The transformation MUST be one of the following:
A single scale or identity transformation
A sequence transformation containing one scale and one translation transformation.

So... Is it a list now, or a single transformation (being either of type scale, identity or sequence)? The current phrasing almost reads like I'm supposed to put a single transform inside a list, and that single transform might be a sequence-type if I wanted both scale and translation.

I assume since all of the allowed cases map to exactly one transformation (which may be a sequence that nests two children), the paragraph above is leftover phrasing from before RFC-5 and should simply say "whose value is an object that defines...".

Or are we doing the list-with-one-entry thing to stay backwards compatible? Doubtful value, since nesting axes inside coordinateSystems is a breaking change anyway, so 0.5 readers won't have any axis information to contextualise the transforms with 🙃

[btbest]

Looking through the examples in this repo, it looks like single-transform-inside-a-list it is...

So then the practical rephrasing suggestion is very minor:

Each object in datasets MUST contain the field coordinateTransformations, whose value is an array containing a single object defining a transformation that maps Zarr array coordinates for this resolution level to the “intrinsic” coordinate system.