Learned-key sensitivity, brute-force and real-token stages; legend order

exp_learned.py completes the learned side of the result stages, so every
figure can carry both realizations of keyed masking. real() holds the
structured artifacts aside and restores them, since exp_real_sec writes
fixed file names.

replot_security.py ranks legend handles from one declared order at all
three ax.legend call sites, so entries no longer follow plot-call order
and drift between figures.

data/sec_brute_learned.csv

@@ -0,0 +1,15 @@
+K,ser_mask
+1,0.9967928683
+3,0.9934662748
+10,0.9869523182
+30,0.9769415244
+100,0.961966217
+300,0.9457608404
+1000,0.9257560018
+3000,0.8881725568
+10000,0.8471349462
+30000,0.8132026814
+65536,0.790006818
+100000,0.7797171991
+300000,0.7435325695
+1000000,0.7188559867