Merge stacked feature branches into main

[ammarcsj]

Background

This work was developed as a stack of 11 PRs (#127–#137), each targeting the
previous branch. After #127 was merged into main, the base of #128 was not
updated to main before merging, so it merged into remove_dead_code instead.
Every subsequent PR then stacked on top of that, with all changes accumulating
on add_quant_reader_and_table_updates. This PR merges that consolidated branch
into main.

Individual PRs (for history and review)

• Remove orphan functions and dead plotting/utility code #127 Remove orphan functions and dead plotting/utility code
• Add graphviz pure-Python fallback layout for treeviz #128 Add graphviz pure-Python fallback layout for treeviz
• Update quant reader config and table handling #129 Update quant reader config and table handling
• Add split ion backgrounds and new z-aggregation modes #130 Add split ion backgrounds and new z-aggregation modes
• Add variance predictor module #131 Add variance predictor module
• Add intensity summarization module #132 Add intensity summarization module
• Add ICC correction module #133 Add ICC correction module
• Add residual decorrelation module #134 Add residual decorrelation module
• Wire new aggregation features into cluster and pipeline #135 Wire new aggregation features into cluster and pipeline
• Add documentation to residual_decorrelation module #136 Add documentation to residual_decorrelation module
• Restore notebook plotting functions removed by dead code cleanup #137 Restore notebook plotting functions removed by dead code cleanup