nangate45: bump finish__timing__setup__tns thresholds for openroad bump

Jenkins CI on PR The-OpenROAD-Project#4208 reports two metric drifts after the
tools/OpenROAD bump in this PR:

- nangate45/bp_fe (base): -1.21838 vs threshold -0.685 -> -1.3
- nangate45/mempool_group (base): -11515.9 vs threshold -11500 -> -11600

Both designs reach the last stage and only finish__timing__setup__tns
drifted. Loosen the thresholds with a small margin so CI passes; the
underlying timing change is from openroad's own optimizer, not this
PR's bazel-orfs work.

Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>

Author: oharboe

flow/designs/nangate45/bp_fe_top/rules-base.json

@@ -84,7 +84,7 @@
         "compare": ">="
     },
     "finish__timing__setup__tns": {
-        "value": -0.685,
+        "value": -1.3,
         "compare": ">="
     },
     "finish__timing__hold__ws": {

flow/designs/nangate45/mempool_group/rules-base.json

@@ -84,7 +84,7 @@
         "compare": ">="
     },
     "finish__timing__setup__tns": {
-        "value": -11500.0,
+        "value": -11600.0,
         "compare": ">="
     },
     "finish__timing__hold__ws": {