Use consistent names for `Decidable` combinators (#2842 redux)

[jamesmckinna]

As #2843 makes clear, this issue proliferates beyond the Relation.Nullary.Decidable.Core combinators, but how to handle that (downstream!) should probably be separate issues (incl. eg. #2845 ), but with callbacks to #2842 ...

UPDATED: ugh. shopping list needs revision, because GitHub isn't linking to the relevant comments correctly?!

Shopping list:

• ≡?: [ fix ] Relation.Nullary.Decidable.Core names for combinators #2843 (comment) FIXED in [ breaking ] change fieldname _≟_ to _≈?_ in IsDecEquivalence #2952
• _∩?_: [ fix ] Relation.Nullary.Decidable.Core names for combinators #2843 (comment) See [ rename ] decidability for Relation.Binary.Construct.{Intersection|Union}` #2955 and for _∪?_ as well
• pointwise?: [ fix ] Relation.Nullary.Decidable.Core names for combinators #2843 (comment) on balance,
  • I (used to) think that 'fixing' this one would be a jarring naming inconsistency wrt all the other combinators defined in that module
  • Pointwise is already arguably a problem wrt discoverability, and adding another cognate name helps sediment that decision; UPDATED so now consider changing the names altogether!
  • UPDATED: see [ refactor ] name of Data.Product.Relation.Binary.Pointwise.NonDependent.×-decidable #2958
• ×? in Data.Product.Relation.Binary.Lex.Strict: [ fix ] Relation.Nullary.Decidable.Core names for combinators #2843 (comment) I think similar conclusions apply as to pointwise?...? So the real question becomes what name to choose for the relation and its decidability, given that Lex takes three arguments
• <ₗₑₓ?: [ fix ] Relation.Nullary.Decidable.Core names for combinators #2843 (comment) Ditto., but worth fixing the assumption names? [ fix ] assumption name in Data.Product.Relation.Binary.Lex.Strict #2956

NB. The general situation for combinators is a bit mixed, both in their definition, and also in their use-sites, which might otherwise require some qualified name/renaming shenanigans to get things to work nicely. There is already an amount of that even in #2952 wrt ≡? vs. ≈? in order to disambiguate lemma statements. Sigh.