Update population size documentation and parameter rationale in ARTnet

[smjenness]

Context

Wrap-up issue. Once the upstream popsize2026 issues close and a decision is reached on whether to keep Grey's 102K, switch to Islek's 53K, use Bennett-derived ~72K, or do something else, ARTnet's documentation needs to be updated to explain the choice and its provenance.

This is intentionally the last issue in the popsize2026 set. It depends on the outcomes of #79, #80, #81, #82, and #83, and should not be started until those produce a decision.

Tasks

• Update package-level documentation explaining the population size parameter, its source, the year of estimate, the recall window, and the behavioral definition behind it.
• Update ARTnet vignettes/examples that use the Atlanta MSM parameterization to reflect the chosen N and rationale.
• Provide a short methods-section blurb (drop-in citation paragraph) for downstream EpiModelHIV publications going forward.
• Update CLAUDE.md if the architectural notes about N are affected.
• Consider: ship a small companion vignette discussing the Grey/Islek/PHSKC trade-offs so future users (and future-us) understand why this N was chosen and not the others.

Dependencies

Depends on:

• Investigate NHANES vs. meta-analytic MSM prevalence: source comparison #79 (NHANES vs meta-analytic prevalence)
• Audit surveillance denominator definitions for MSM HIV/PrEP rates #80 (surveillance denominator audit)
• Document PHSKC / BRFSS-based MSM prevalence as third reference point #81 (PHSKC/BRFSS reference)
• Explore within-individual variation in MSM sexual activity over time using ARTnet partnership data #82 (within-individual activity variation in ARTnet)
• Sensitivity check: regenerate target statistics at multiple N values #83 (target-stat sensitivity to N)

Priority

Low. Wrap-up only.

Source

Plan: inst/popsize/msm_population_size_discussion.md, §Q6.