\begin{table}[p]
\centering\scriptsize
\begin{threeparttable}
\caption{Monte Carlo, exp contam design, $p = 1$, $N = 1,048,576$: scaled gap to $\widehat\theta_N$, scaled RMSE and 95\% coverage by local sample size $m$.}\label{tab:mc_exp_contam_p1}
\begin{tabular}{@{}lrrrrrrrrr@{}}
\toprule
Estimator & \multicolumn{3}{c}{$m = 16$, $K = 65,536$} & \multicolumn{3}{c}{$m = 64$, $K = 16,384$} & \multicolumn{3}{c}{$m = 256$, $K = 4,096$} \\
 & gap & RMSE & cov & gap & RMSE & cov & gap & RMSE & cov \\
\midrule
(1,0) plain average & 101.734 & 17.256 & 0.000 & 30.226 & 54.426 & 0.000 & 8.279 & 76.236 & 0.000 \\
(1,1) average + 1 Newton & 11.018 & 95.547 & 0.000 & 0.973 & 85.618 & 0.000 & 0.073 & 84.585 & 0.000 \\
(1,2) average + 2 Newton & 0.129 & 84.659 & 0.000 & 0.001 & 84.646 & 0.000 & 5.71e-6 & 84.512 & 0.000 \\
(2,0) half-block jackknife & 12.900 & 71.658 & 0.000 & 4.719 & 79.946 & 0.000 & 0.607 & 83.956 & 0.000 \\
(2,1) half-block + 1 Newton & 0.179 & 84.707 & 0.000 & 0.024 & 84.669 & 0.000 & 4.98e-4 & 84.513 & 0.000 \\
(2,2) half-block + 2 Newton & 3.51e-5 & 84.530 & -- & 6.55e-7 & 84.645 & -- & 5.03e-10 & 84.512 & -- \\
(3,0) quarter-block & 21.466 & 63.100 & -- & 3.068 & 81.628 & -- & 0.361 & 84.360 & -- \\
(3,1) quarter-block + 1 Newton & 0.493 & 85.020 & 0.000 & 0.011 & 84.655 & 0.000 & 2.37e-4 & 84.512 & 0.000 \\
analytic Cox--Snell + 1 Newton & 1.432 & 85.961 & 0.000 & 0.242 & 84.888 & 0.000 & 0.022 & 84.535 & 0.000 \\
CSL, 1 round & 1.39e-13 & 84.530 & -- & 3.13e-13 & 84.645 & -- & 3.19e-13 & 84.512 & -- \\
CSL, 2 rounds & 1.35e-13 & 84.530 & -- & 3.12e-13 & 84.645 & -- & 3.16e-13 & 84.512 & -- \\
AEE (Lin--Xi) & 120.265 & 204.789 & -- & 33.285 & 117.926 & -- & 8.542 & 93.052 & -- \\
(2,1) stale Hessian & 1.806 & 82.728 & -- & 0.191 & 84.456 & -- & 0.006 & 84.507 & -- \\
centralized MLE & 0 & 84.530 & 0.000 & 0 & 84.645 & 0.000 & 0 & 84.512 & 0.000 \\
\bottomrule
\end{tabular}
\begin{tablenotes}\footnotesize
\item 1000 paired replications per column. gap $=$ root mean square of $\sqrt N\,\lVert\tilde\theta - \widehat\theta_N\rVert$; RMSE $=$ root mean square of $\sqrt N\,\lVert\tilde\theta - \theta_0\rVert$; cov $=$ empirical coverage of nominal 95\% Wald intervals (observed information at the estimate; `--' when not computed). Monte Carlo standard errors of the gaps and RMSEs are in \texttt{results/processed/summary.csv}; coverage standard error $\approx 0.007$.
\item Fraction of workers with all local fits valid (orders 2 / 3): $m=16$: 1.000 / 1.000; $m=64$: 1.000 / 1.000; $m=256$: 1.000 / 1.000.
\end{tablenotes}
\end{threeparttable}
\end{table}