1. I am not sure which Jmeter reports you are talking about coz there are lots of plugins available. I can see Gatling gathers very limited amount of data, how many requests are made per any given time of the execution, are the responses OK or KO, what time each request and response take. All this information is stored into simulation.log file. Whereas Jmeter has logs of plugins to get more detailed information. Having said that, the stats reported by both tools should be identical as long as the test setup is similar.
2. Requests per second graph shows how the load is being pushed on the server. Response per second indicates how the server responded to those requests i.e. total, success and failures
If there are zero failures in the test, both graphs should look identical. I hope I put it correctly :)
Note that the requests per second graph corresponds to the ACTUAL number requests triggered by the script, it is not necessarily the number of transactions that you would have defined in the script. I know this statement is a bit confusing. So example comes to rescue.
startTransaction
httpRequest1
httpRequest2
httpRequest3
EndTransaction
One execution of above will show 3 requests in request per second graph, unlike tTPS graph in Loadrunner which will show only 1 request (for your transaction). This is what I observed.
@Stephane, thoughts?