Implementing a cache mechanism like Memcached or Redis is the correct approach. It directly reduces datastore reads, thereby lowering latency during peak hours. Automatic scaling is not an option for the local development server, manual scaling is inefficient due to the need for constant monitoring, and restarting the service via cron jobs does not address the root cause of latency issues.