Background/Objectives: Total neoadjuvant therapy (TNT) is standard for locally advanced rectal cancer, yet treatment response is biologically heterogeneous, underscoring the need for predictive biomarkers. This study aimed to characterize how baseline and on-treatment gut microbiome ecology relate to clinically relevant response dimensions, and to identify baseline-level markers that could be integrated into a pre-treatment response score. Methods: Shotgun metagenomic profiling of the stool microbiome was performed in a real-world cohort of 13 patients with locally advanced rectal cancer treated with TNT followed by total mesorectal excision, together with 18 non-RC controls. Treatment response was assessed in a pathology-anchored framework integrating both tumour regression and nodal response, with favourable response defined by Modified Ryan TRG1–2 and two-category nodal downstaging. Results: Longitudinal analysis from the therapy-naive baseline through TNT to the end of treatment showed that tumour topography was associated with distinct baseline microbiome states and divergent treatment-associated remodelling trajectories. In our cohort, age also stratified microbiome behaviour, with patients aged ≥70 years showing higher baseline pathogen-associated and antimicrobial resistance-linked burden together with a more contraction-prone ecological trajectory. Partitioning the microbiome into below-mean and above-mean fractions showed that biomass was concentrated in dominant taxa, whereas much of the diversity resided in the low-abundance background. Favourable TNT response was associated more with preservation of this low-abundance complexity than with higher overall diversity alone. TNT remodelled, but did not eliminate, a microbiome layer enriched in genotoxicity- and virulence-associated signals, which remained structured by age, tumour location, and response category. Among baseline taxa, Phocaeicola coprophilus emerged as the only species-level signal consistently associated with favourable outcome across both pathological and nodal response dimensions, discriminating patients with concordantly favourable outcomes from all others. To integrate the most informative pre-treatment microbiome features, an exploratory Microbiome Baseline Score was generated, showing promising preliminary performance in this pilot cohort, with a sensitivity of 0.875 and a specificity of 1.00, warranting validation in larger independent cohorts. Conclusions: Our results support the concept that clinically relevant response-associated information may already be embedded in the therapy-naive gut microbiome and can be condensed into a proof-of-concept baseline score for future microbiome-informed pre-treatment stratification in LARC.