Check test result via an assertion

As long as `has_no_branch?` is only called in an assertion it
shouldn't matter if it could return false but still allow the test to
continue.

So we don't need the new wait method

qa/qa/page/project/branches/show.rb

@@ -29,7 +29,7 @@ module QA
 
           def has_no_branch?(branch_name)
             within_element(:all_branches) do
-              has_no_css?(".js-branch-#{branch_name}")
+              has_no_css?(".js-branch-#{branch_name}", wait: Support::Waiter::DEFAULT_MAX_WAIT_TIME)
             end
           end
 
@@ -46,14 +46,6 @@ module QA
               click_element(:delete_merged_branches)
             end
           end
-
-          def wait_for_branch_not_present(branch_name)
-            branch_not_present = wait(reload: false) do
-              has_no_branch?(branch_name)
-            end
-
-            raise "Expected branch `#{branch_name}` not to be present" unless branch_not_present
-          end
         end
       end
     end

qa/qa/specs/features/browser_ui/3_create/repository/add_list_delete_branches_spec.rb

@@ -72,7 +72,7 @@ module QA
 
         Page::Project::Branches::Show.perform do |branches_view|
           branches_view.delete_branch(third_branch)
-          branches_view.wait_for_branch_not_present(third_branch)
+          expect(branches_view).to have_no_branch(third_branch)
         end
 
         Page::Project::Branches::Show.perform(&:delete_merged_branches)
@@ -83,7 +83,7 @@ module QA
 
         page.refresh
         Page::Project::Branches::Show.perform do |branches_view|
-          branches_view.wait_for_branch_not_present(second_branch)
+          expect(branches_view).to have_no_branch(second_branch)
         end
       end
     end