Inconsistent Burr XII key: 'burr' vs 'burr12' across pipelines

[cm401]

Summary

The codebase uses two different string keys for the Burr XII distribution depending on which pipeline is involved:

• Meta-analysis pipeline (main_analysis.R, run_meta_analysis.R, table_utils.R, make_supplementary_figures.R): uses "burr"
• Simulation pipeline (utils.R, new_simulation_study.R, run_simulation_study.R): uses "burr12"

Current impact

No active bug: the meta-analysis pipeline is self-consistent, so dist_labels["burr"] resolves correctly when looking up stan_best_dist. However, this is a latent risk if:

• model_weights.rds is ever regenerated using "burr12" keys
• The two pipelines are unified or share data structures

Suggested fix

Standardise on a single key throughout the codebase. "burr12" is the more precise identifier and is already used in the simulation pipeline and Stan model code. The meta-analysis pipeline should be updated to match.

As a short-term defensive measure, dist_labels in run_meta_analysis.R could include both keys:

dist_labels <- c(
  lognormal = "Log-normal",
  gamma     = "Gamma",
  weibull   = "Weibull",
  burr      = "Burr XII",
  burr12    = "Burr XII",
  gengamma  = "Gen. Gamma"
)