Update Scalafmt to 2.4.2

[travisbrown]

This is a more conservative version of #3315 that still rewrites { } to ( ) for single-line expressions, but doesn't remove newlines in cases like this, which was my primary complaint about the default config in 2.4.2:

  forAll { (nes: NonEmptySet[Int]) =>
    nes.reduce should ===(nes.fold)
  }

In #3315 that got rewritten to the following, which obscures what's being tested, at least to my eyes:

  forAll((nes: NonEmptySet[Int]) => nes.reduce should ===(nes.fold))

Instead of +495 −1,432 it's +58, −50, so quite a bit less dramatic, and the changes look more reasonable to me.

[codecov-io]

Codecov Report

Merging #3337 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #3337      +/-   ##
==========================================
+ Coverage   93.31%   93.31%   +<.01%     
==========================================
  Files         378      378              
  Lines        7688     7689       +1     
  Branches      206      206              
==========================================
+ Hits         7174     7175       +1     
  Misses        514      514

Flag	Coverage Δ
#scala_version_212	93.38% <100%> (ø)	⬆️
#scala_version_213	93.09% <100%> (ø)	⬆️

Impacted Files	Coverage Δ
core/src/main/scala/cats/syntax/monadError.scala	80% <100%> (+2.22%)	⬆️
core/src/main/scala/cats/data/WriterT.scala	91.26% <100%> (ø)	⬆️
...eycats-core/src/main/scala/alleycats/std/map.scala	46.42% <100%> (ø)	⬆️
core/src/main/scala/cats/data/Func.scala	96.42% <100%> (ø)	⬆️
core/src/main/scala/cats/data/Cokleisli.scala	100% <100%> (ø)	⬆️
core/src/main/scala/cats/data/Kleisli.scala	90.82% <100%> (ø)	⬆️
core/src/main/scala/cats/instances/map.scala	95.91% <100%> (ø)	⬆️
core/src/main/scala/cats/instances/sortedMap.scala	86.15% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 482660e...81b7c50. Read the comment docs.

[kailuowang]

I prefer this format as well

[rossabaker]

I mostly don't like deviating from the defaults, but I'm persuaded on this one.

[travisbrown]

Thanks all! We can always reconsider going back to defaults later.