A source coefficient needs its actual comparison map

The supplied Em02-A record calculates localization powers in $ϵ$ for its displayed scalar seeds CheungEm02A. Em02-D calculates an inverse-cubic smoothing coefficient for its fixed distribution model CheungEm02D. Neither calculation defines a relational-lapse quantum contraction or an incidence-register kernel.

The parameters also remain different:

tends to zero along $ϵ={\delta }^2$, is constant along $ϵ={\delta }^{3/2}$, and diverges along $ϵ=\delta$. A joint limit needs its own regime.

An $O({\delta }^{-2})$ remainder after a leading ${\delta }^{-3}$ term is neither a finite part nor an evaluated remainder divergence. Both zero and ${\delta }^{-2}C$ satisfy such an upper-bound shape. No lower-order subtraction for another observable is licensed by it.