\varinstantiate and sequences.

[kohlhase]

In smglom/topology/source/producttopology.en.tex we have a case, where we need a (finite) sequence of math structures that would normally be declared by \varinstantiate. It seems that the sequences mechanism is orthogonal to all other declaration mechamisms, and that (probably) all combinations will eventually occur. Thus we need to design a general mechanism and surface syntax for that.

[Jazzpirate]

Entails:

• various kinds of declarations should be "sequentializable". \[var]instantiate ist just one instance (pun intended) of this problem. Ideally, \varseq would be paramtric in an arbitrary declaration constructor. Not sure if that's feasible. Might entail completely rewriting core mechanisms.
• every field in every instance in a sequence needs a notation
• also in expectation for conservative extentions for {mathstructure}, \instantiate and \varseq should probably be dynamic in the implementation, rather than (as currently) generating macros for every field at point of declaration. That might also help with the orthogonality problem.