If you talk about the results vs performances, both are used for entirely different purpose.
With JMeter, you can speedup the communication with the server and so you can make sure that specific function works fine or not based on the response. But, it does not really gaurrentee you that your UI works as expected.
Selenium simply helps you imitates user actions on the browser of your choice. So, you can write your code once and execute it on different set of browsers to make sure that all functionality works fine on all browser.
I will give you an example here.
Say, you have an application which has many Ajax calls. Now, you have a change password functionality that changes password through AJAX call. It is very important for you to verify that this AJAX works across all different browsers. So, if you would have used Selenium, you can easily say that your JavaScript code works fine.
But, using JMeter, it might complete the AJAX call. But, it does not say that your JavaScript function works across all browsers.
Hope I am clear.