More "intelligence" in functionals?

[kohr-h]

This is a meta-issue for a bunch of things that would add more intelligence in the functionals.

• Should scaling and addition of SeparableSum yield a SeparableSum? indexing for separable sums #897 No, see discussion.
• Add strong convexity constant with corresponding calculus? Strong convexity #1078
• Add Bregman divergence? Bregman distances for functionals #918 Done in Issue 918 bregman functional #1267