Merging: mirror the negative-intensity Poisson guard on the GPU

6be94f2be stopped subtracting a negative intensity's Poisson term from the
background variance, but only in the host Combine(). The CUDA combine is the
path that actually runs: Run() selects it whenever a device is present and no
observation dump was asked for, so the correction never took effect on a normal
run, and a --dump-observations run merged differently from a normal one - the
two are meant to be identical.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

image_analysis/scale_merge/RotationScaleMergeGPU.cu

@@ -354,7 +354,9 @@ namespace {
                     const double corr = p.corr[i];
                     const double I_corr = pooled_I(i) * corr;
                     const double sigma_corr = double(p.sigma[i]) * corr;
-                    const double bkg_var = sigma_corr * sigma_corr - corr * I_corr;
+                    // max(0, I): a down-fluctuated partial carries no Poisson signal to remove, and
+                    // removing a negative one inflates the background part instead of leaving it alone.
+                    const double bkg_var = sigma_corr * sigma_corr - corr * Dmax(0.0, I_corr);
                     double var = Dmax(0.0, bkg_var) + corr * Dmax(0.0, F);
                     if (!(var > 0.0)) var = sigma_corr * sigma_corr;
                     const double w = 1.0 / var;