small cleanup of subducting plate

[MFraters]

Mostly to test the new benchmark introduced in #266, but the changes clean the code up as well.

[coveralls]

Pull Request Test Coverage Report for Build 940497608

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage remained the same at 98.535%

---

Totals
Change from base Build 940479654:	0.0%
Covered Lines:	4573
Relevant Lines:	4641

---

💛 - Coveralls

[codecov]

Codecov Report

Merging #267 (be30c1d) into master (092c216) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master     #267   +/-   ##
=======================================
  Coverage   98.45%   98.45%           
=======================================
  Files          74       74           
  Lines        4397     4397           
=======================================
  Hits         4329     4329           
  Misses         68       68           

Impacted Files	Coverage Δ
source/features/subducting_plate.cc	99.09% <ø> (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 092c216...be30c1d. Read the comment docs.

[github-actions]

Benchmark	Master	Feature	Difference (99.9% CI)
Slab interpolation simple none	2.944 ± 0.050 (s=154)	2.948 ± 0.046 (s=154)	-0.5% .. +0.8%
Slab interpolation curved simple none	3.078 ± 0.065 (s=160)	3.065 ± 0.066 (s=136)	-1.3% .. +0.4%
Spherical slab interpolation simple none	2.694 ± 0.150 (s=161)	2.700 ± 0.178 (s=175)	-2.0% .. +2.4%
Slab interpolation simple curved CMS	3.453 ± 0.115 (s=133)	3.449 ± 0.102 (s=130)	-1.4% .. +1.2%
Spherical slab interpolation simple CMS	8.294 ± 0.191 (s=51)	8.327 ± 0.159 (s=60)	-1.0% .. +1.8%
Spherical fault interpolation simple none	5.462 ± 0.142 (s=74)	5.493 ± 0.161 (s=93)	-0.9% .. +2.0%

[MFraters]

It is clear that the accuracy of the slower tests has gone down a bit due to less runs in the 15 minute benchmark time. So I hope #264 brings the run time down again or otherwise the tests has to be simplified a bit or extend the run time.