I don't think it's a good idea to allow the arguments block warning to be disabled.  Like the warning says, incorrect results are possible if execution of arguments blocks are simply skipped.  It seems bad to me to allow that without a warning.

I suppose another option is to just disable the parsing of arguments blocks until they are completely implemented so you get an unconditional error instead of an unconditional warning.

There were a couple other warnings in the same file without ids, so for good measure I changed all of them as follows. Does this change work for you?

diff -r fd5ae4140483 libinterp/parse-tree/pt-eval.cc
--- a/libinterp/parse-tree/pt-eval.cc   Wed Oct 18 17:04:46 2023 -0400
+++ b/libinterp/parse-tree/pt-eval.cc   Wed Oct 18 17:26:47 2023 -0400
@@ -1257,7 +1257,8 @@ tree_evaluator::visit_argument_list (tre
 void
 tree_evaluator::visit_arguments_block (tree_arguments_block&)
 {
-  warning ("function arguments validation blocks are not supported; INCORRECT RESULTS ARE POSSIBLE");
+  warning_with_id ("Octave:function-arg-blocks-unsupported",
+  "function arguments validation blocks are not supported; INCORRECT RESULTS ARE POSSIBLE");
 }

 void
@@ -2702,7 +2703,8 @@ void tree_evaluator::mlock (bool skip_fi

   if (fcn->is_builtin_function ())
     {
-      warning ("mlock: locking built-in function has no effect");
+      warning_with_id ("Octave:lock-built-in-no-effect",
+                       "mlock: locking built-in function has no effect");
       return;
     }

@@ -2718,7 +2720,8 @@ void tree_evaluator::munlock (bool skip_

   if (fcn->is_builtin_function ())
     {
-      warning ("munlock: unlocking built-in function has no effect");
+      warning_with_id ("Octave:unlock-built-in-no-effect",
+                       "munlock: unlocking built-in function has no effect");
       return;
     }