Parametrize CachingOptimizer and MockOptimizer on model type

[blegat]

Initial benchmark:

v0.6
126.218 ns (7 allocations: 176 bytes)
452.726 ns (17 allocations: 416 bytes)
 23.927 ns (3 allocations: 96 bytes)
 40.416 ns (5 allocations: 144 bytes)
v0.7
105.262 ns (5 allocations: 128 bytes)
414.565 ns (12 allocations: 304 bytes)
  9.623 ns (1 allocation: 48 bytes)
 29.113 ns (3 allocations: 96 bytes)

After parametrization of the MockOptimizer:

v0.6
121.149 ns (7 allocations: 176 bytes)
357.014 ns (14 allocations: 352 bytes)
 24.722 ns (3 allocations: 96 bytes)
 24.712 ns (3 allocations: 96 bytes)
v0.7
104.708 ns (5 allocations: 128 bytes)
311.206 ns (9 allocations: 240 bytes)
  9.306 ns (1 allocation: 48 bytes)
  9.400 ns (1 allocation: 48 bytes)

After parametrization of CachingOptimizer

v0.6
 38.889 ns (4 allocations: 112 bytes)
252.048 ns (11 allocations: 288 bytes)
 24.412 ns (3 allocations: 96 bytes)
 24.982 ns (3 allocations: 96 bytes)
v0.7
 23.149 ns (2 allocations: 64 bytes)
207.150 ns (6 allocations: 176 bytes)
  9.761 ns (1 allocation: 48 bytes)
  9.591 ns (1 allocation: 48 bytes)

Related to #321

[codecov-io]

Codecov Report

Merging #323 into master will not change coverage.
The diff coverage is 100%.

@@           Coverage Diff           @@
##           master     #323   +/-   ##
=======================================
  Coverage   96.48%   96.48%           
=======================================
  Files          30       30           
  Lines        4525     4525           
=======================================
  Hits         4366     4366           
  Misses        159      159

Impacted Files	Coverage Δ
src/Utilities/mockoptimizer.jl	96.15% <100%> (ø)	⬆️
src/Utilities/cachingoptimizer.jl	88.71% <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 df5aaaa...366089e. Read the comment docs.

[mlubin]

These numbers aren't particularly helpful given that nobody else has access to the machine where this was timed. I think the link to the issue is sufficient.