feat: Add pointwise conditional log-likelihoods for additional distributions

[sethaxen]

This PR adds support for product distributions, mixture distributions, reshaped distributions, and distributions of joint order statistics. While some of these cases are trivial and others are unlikely to be needed, they make it so that a dependant package can call pointwise_conditional_loglikelihoods on just about anything a user might try to use as an observation model and it will just work. ™️

Motivating example is TuringLang/DynamicPPL.jl#1038

Checklist

• PR follows official PR format
• The new feature is properly documented with an example.
• New or updated tests cover the new feature.
• Code style is correct (formatted using JuliaFormatter with BlueStyle).
• The new feature is listed in the Features section of the changelog if the change is non-breaking or the BREAKING CHANGES section if it is breaking.
• The patch version number been incremented if the change is non-breaking or the minor version number been incremented if the change is breaking.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.31%. Comparing base (ccaccea) to head (705d9f1).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #89      +/-   ##
==========================================
+ Coverage   97.04%   97.31%   +0.27%     
==========================================
  Files          16       16              
  Lines         845      930      +85     
==========================================
+ Hits          820      905      +85     
  Misses         25       25              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.