Prologue to Part II: From "Curve Fitting" to "Data Auditing”
“当解决一个旧问题需要引入更多、更复杂的假设时，往往并不是答案错了，而是问题本身立足的底层逻辑已经失效。”"
When solving an old problem requires introducing more and increasingly complex assumptions, it is often not the answer that is wrong, but the underlying logic upon which the problem stands that has failed.”
Why MOND’s "Strongholds" May Not Be Physical Reality
在第一部分中，我们已经看到，ACM 可以在统一单基准（Single-Baseline）框架下，不依赖逐星系调参，就完成对 SPARC 164 个对比星系的整体拟合。这说明它不仅能把曲线拟合得更好，更重要的是，它提供了一条刚性、统一、可复现的动力学基线。
In Part I, we demonstrated that the ACM framework, operating under a unified Single-Baseline, can achieve a global fit across 164 SPARC galaxies without the need for per-galaxy parameter tuning. This proves more than just superior fitting; it establishes a rigid, unified, and reproducible dynamical baseline.
但问题并没有到此结束。
But the inquiry does not end there.
在主流讨论中，仍然有约 60 多个样本经常被视为 MOND 的“优势区”。这些星系似乎构成了一个顽固的少数派，让 MOND 看起来仍然保留着某种不可替代的解释权。
In mainstream discourse, approximately 60 samples are frequently cited as the "Strongholds" of MOND. These galaxies appear to form a stubborn minority, allowing MOND to retain a seemingly irreplaceable explanatory authority.
可真正值得追问的是：这种“优势”究竟来自物理本身，还是来自观测输入、质量归一化与结构分辨率中的历史局限性？
The critical question, however, remains: Does this "advantage" stem from the underlying physics, or is it a byproduct of historical limitations in observational input, mass normalization, and structural resolution?
这正是 Paper II 要回答的问题。我们不再停留于“哪条曲线更贴数据”，而是进一步对这些样本本身展开一次客观审计。结果显示，那些看似统一的 MOND 优势区，其实并不是一个连贯的物理阵地，而是由三类彼此不同的残差机制拼接而成：
This is the central question of Paper II. We are moving beyond the debate of "which curve fits better" to conduct an objective Forensic Audit of the samples themselves. The results reveal that these alleged MOND strongholds do not constitute a coherent physical front. Instead, they are a patchwork of three distinct residual mechanisms: