Pretty-printing binops should have lower affinity somehow

[graydon]

When printing a binop expression containing fn calls, we wind up wrapping rather late:

foo(a, b, c) + bar(d, e, f)

will sometimes print as

foo(a, b, c) + bar(d, e,
                   f)

rather than

foo(a, b, c)
+ bar(d, e, f)

or similar (+ could be on the previous line, whatever). It seems like maybe extra boxes or higher-precedence boxes or something should enclose the fn calls?

[msullivan]

Graydon says this is probably fixed.