Your best bet for now is to disable the test in these cases. The original test was (very) wrong and the new test is incomplete and thus broken for now. The new code could look something like:

   my @req_prm = $stmt->params ();
   unless (@req_prm == 1 && ref $req_prm[0] eq "SQL::Statement::Param"
+) {
       my $n_req = @req_prm == 1 && ref $req_prm[0] ? $req_prm[0]->num
+ : scalar @req_prm;
       unless ($n_req == (my $nparm = @$params)) {
           my $msg = "You passed $nparm parameters where $n_req requir
+ed";
           $sth->set_err ($DBI::stderr, $msg);
           return;
           }
       }
[download]

But do not see this as a fix, as it is cowardly running away from problems hiding in the dark.