Sounds too good to be true, right?
Probably one reason for free version is that as Apache Jmeter is de facto standard for Web applications because it is free, working well in most cases and widely used. Usually Jmeter is default option for running performance tests until there are special requirements advising to use other tools.
However if you are ok with 50 concurrent users choice is not so obvious any more as NeoLoad and LoadRunner are actually better tools compared to Jmeter and have advanced features Jmeter is missing.
In practise with simple Web applications with Continuous need for testing Jmeter is safer choice but there are also cases where NeoLoad and LoadRunner are better choices like Applications with lots of dynamic correlation.
With free version it is easy to try these tools and obviously decision to use them also in bigger cases and pay license fees is easy once tool is verified against Application to be tested.
These Free tools have offered welcome options and for short projects risks are small. Obviously these tools are not Free maybe forever but until this they are good options for any new Performance testing case.