feat: investigate maxk differences and add galsim algorithm for maxk

[beckermr]

I have developed a scan-based implementation of galsim's algorithm for maxk. This algorithm is ~10-20% slower than the one we use now. In most cases it does not make a difference which one we use, so I am opting for the faster one.

I also went through some of the changes in the test suite for maxk and adjusted the threshold for one test for Moffat profiles. I also enabled more of the Moffat tests to be run.

closes #225
closes #189

For associated changes to the test suite see: GalSim-developers/GalSim-for-JAX-GalSim-Testing#4

[codspeed-hq]

Merging this PR will degrade performance by 32.22%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 7 improved benchmarks
❌ 1 regressed benchmark
✅ 28 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	WallTime	test_benchmarks_lanczos_interp[xval-no_conserve_dc-run]	114.1 µs	72.8 µs	+56.76%
⚡	WallTime	test_benchmarks_lanczos_interp[xval-conserve_dc-run]	129.5 µs	81.2 µs	+59.61%
⚡	WallTime	test_benchmarks_lanczos_interp[kval-no_conserve_dc-run]	56.4 µs	41.3 µs	+36.75%
⚡	WallTime	test_benchmarks_lanczos_interp[kval-conserve_dc-run]	55.3 µs	40.6 µs	+36.33%
⚡	WallTime	test_benchmark_moffat_init[run]	97.3 µs	67.7 µs	+43.65%
⚡	WallTime	test_benchmark_spergel_calcfluxrad[run]	215.2 µs	165.4 µs	+30.07%
⚡	WallTime	test_benchmark_invert_ab_noraise[run]	144.7 µs	104.2 µs	+38.85%
❌	WallTime	test_benchmark_spergel_conv[run]	156.4 ms	230.7 ms	-32.22%

---

_{Comparing fix-maxk (ffcef75) with main (e3ff71f)}

[beckermr]

OK This Pr is ready to go. It documents a bunch of work I've done to investigate how we set maxk for interpolated images. TL;DR; what we do now matches galsim exactly in most situations. We can write an algorithm that does exactly what galsim does, but it is more expensive and so doesn't seem worth the trouble. I have left the new algorithm commented out in the code. We can switch to it in the future if we feel it is important.

[beckermr]

@ismael-mendoza this one is ready to go!

[ismael-mendoza]

LGTM!