Switch SCS timings to seconds

[SteveDiamond]

Description

Switches SCS timings from milliseconds to seconds.
Issue link (if applicable): #1878

Type of change

• New feature (backwards compatible)
• New feature (breaking API changes)
• Bug fix
• Other (Documentation, CI, ...)

Contribution checklist

• Add our license to new files.
• Check that your code adheres to our coding style.
• Write unittests.
• Run the unittests and check that they’re passing.
• Run the benchmarks to make sure your change doesn’t introduce a regression.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

Steven Diamond seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[github-actions]

Benchmarks that have stayed the same:

   before           after         ratio
 [129ea157]       [45d4f612]
      32.4±0s          33.3±0s     1.03  cvar_benchmark.CVaRBenchmark.time_compile_problem
      1.98±0s          2.02±0s     1.02  simple_LP_benchmarks.SimpleScalarParametrizedLPBenchmark.time_compile_problem
      10.2±0s          10.3±0s     1.01  semidefinite_programming.SemidefiniteProgramming.time_compile_problem
      48.6±0s          48.7±0s     1.00  sdp_segfault_1132_benchmark.SDPSegfault1132Benchmark.time_compile_problem
      9.92±0s          9.95±0s     1.00  optimal_advertising.OptimalAdvertising.time_compile_problem
      4.36±0s          4.35±0s     1.00  simple_LP_benchmarks.SimpleFullyParametrizedLPBenchmark.time_compile_problem
      26.4±0s          26.2±0s     0.99  qp_1611_benchmark.QP1611Benchmark.time_compile_problem
      26.3±0s          26.0±0s     0.99  simple_LP_benchmarks.SimpleLPBenchmark.time_compile_problem
      12.4±0s          12.3±0s     0.99  svm_l1_regularization.SVMWithL1Regularization.time_compile_problem
      10.7±0s          10.1±0s     0.95  huber_regression.HuberRegression.time_compile_problem

[SteveDiamond]

Fixes #1878. Failures unrelated and known mosek tests.

[phschiele]

Confirming that SCS 2.x and 3.x are affected:
2.x:

63.798962 583.526313

3.x:

92.963834 2936.187088

Side note: Interesting to see that 2.x was much faster on this problem instance (I fixed eps=1e-4 in both cases).