The biggest problems with this solution have to do with scale. If the load is high then your request is put on hold for a second, and then you have to handle it. If you make a lot of requests then it is your problem. It would be very helpful if a queue was implemented to handle, for example, 100 requests at the same time. Any additional request would be put on hold and made to wait for a few seconds. Once the network and infrastructure are loaded to handle the next request, it would proceed.