Improvements to make for code quality

- Clarify in the HTML report what determines a "valid" image pair

- Have 2 separate thresholds for R and U success? Put in place right before the results against the metric
- Could add image-only metrics before Verifier Summary

- Make a single table for each module (instead of FrontEnd Summary, have Verifier Summary, and PostProcessor Summary)
- There should be a #avg number of inliers w.r.t. est model for Veriifer, and also separately for the PostProcessor
- Create PostProcessor class in accordance with the graph

- Decide upon what goes in side-by-side plots when comparing with ground truth
- Should COLMAP # tracks go into COLMAP column? or should it be error w.r.t. COLMAP?

No file, test, or entry point is named. Begin by tracing the HTML report sections, including Verifier Summary, PostProcessor Summary, image-only metrics, and side-by-side ground-truth plots, then review the linked graph for the proposed PostProcessor class. Done requires resolving the listed metric, threshold, table, and plot decisions and implementing the agreed report structure.

Written by the indexing model from the issue text.