The following posts deal with performance problems in similiar scenarios:
https://discourse.geta12.com/t/how-to-not-block-resources-while-waiting-for-solr-commits/2493
https://discourse.geta12.com/t/how-to-improve-overall-dataservices-performance-because-of-poor-hazelcast-caching/2659/2
You might want to have a look and try to implement the changes described there.
Doing more requests in parallel won’t probably help anything because all the processing will wait for solr hard commits after every document update. You will just be blocking more servlet request processing threads by using batching in frontend.