This topic may be helpful to find out what parts are slow.
Also look at use of system resources to see if there are any bottlenecks. It seems you still have a max concurrent requests limit set, otherwise you would have no rejected transfers. What did you set it to?
The issue linked by @Cmdrd will also have a big impact though. So looking at other things should be done in the context of trying to optimize, not because there is necessarily anything wrong with your node.