Improved grunt express/casper watchers

closes TryGhost#9718

- fs.fileWatch that is used internally by 'gase' in 'grunt-contrib-watch', is having 100ms pooling default (https://github.com/shama/gaze/blob/07828a684566b6d4844f12b747e74e376fa31744/lib/gaze.js#L36). This is causing hight CPU usage for large amount of files.
- As suggested in https://github.com/gruntjs/grunt-contrib-watch#why-is-the-watch-devouring-all-my-memorycpu the watch interval was set to higher 500ms because the recommended default of 5s (gruntjs/grunt-contrib-watch#145 (comment)) was visible in the development flow

Gruntfile.js

@@ -70,7 +70,8 @@ const configureGrunt = function (grunt) {
                     'content/themes/casper/assets/js/*.js'
                 ],
                 options: {
-                    livereload: true
+                    livereload: true,
+                    interval: 500
                 }
             },
             express: {
@@ -84,7 +85,8 @@ const configureGrunt = function (grunt) {
                 tasks: ['express:dev'],
                 options: {
                     spawn: false,
-                    livereload: true
+                    livereload: true,
+                    interval: 500
                 }
             }
         },