I cannot vouch for it's comparison with Jmeter, but I have used TC for load tests and stress tests of web applications. In my use of it, we uncovered a number of bottlenecks in our application, tested out prototypes of solutions for those same bottlenecks, and evaluated "crash" conditions to determine the upper limit of load for our sites.
Is it better or worse than Jmeter? I think when it comes to any kind of testing, the tool is only as good as the people who use it. I know Amish men who can put a roof on a house faster with a hand hammer than a professional contractor can with a air-powered nail gun. It's how skilled you are in the style of test that makes the biggest difference.