fix NaN assertions that never fired due to IEEE 754 NaN comparison

[chongchonghe]

Description

Fix debug assertions that silently passed even when variables held NaN values, due to incorrect use of relational operators with NAN.

In IEEE 754, NaN != NaN always evaluates to true, so AMREX_ASSERT(x != NAN) and AMREX_ASSERT_WITH_MESSAGE(x != NAN, ...) are no-ops — they never fire regardless of whether x is NaN. Two sites were affected:

• RadSystem::ComputeRadPressure in src/radiation/radiation_system.hpp: four assertions on Fn, Tnx, Tny, Tnz that guard against uninitialized flux/pressure-tensor components.
• problem_main in src/problems/ParticleSink/testParticleSink.cpp: one assertion that boost_vel_x was actually read from the input file.

Both sites are replaced with std::isfinite(value), which correctly returns false for NaN and infinity.

Related issues

Fixes #1857.

Checklist

Before this pull request can be reviewed, all of these tasks should be completed. Denote completed tasks with an x inside the square brackets [ ] in the Markdown source below:

• I have added a description (see above).
• I have added a link to any related issues (if applicable; see above).
• I have read the Contributing Guide.
• I have added tests for any new physics that this PR adds to the code.
• (For quokka-astro org members) I have manually triggered the GPU tests with the magic comment /azp run.

[gemini-code-assist]

Code Review

This pull request replaces incorrect != NAN comparisons with std::isfinite() checks in testParticleSink.cpp and radiation_system.hpp to properly handle NaN values. A review comment suggests using AMREX_ALWAYS_ASSERT_WITH_MESSAGE for input validation in testParticleSink.cpp to ensure the check remains active in release builds.