If you want the raw response string out of SolrJ, the advice seems to be to just use an HttpClient directly.
However, sometimes you may have a lot of SolrJ infrastructure already in place to build out queries, etc, so it would be much simpler to just use SolrJ to do the work.
This patch offers a NoOpResponseParser, which simply puts the entire response into an entry in a NamedList.
Because the response isn't parsed into a QueryResponse, usage is slightly different:
HttpSolrServer server = new HttpSolrServer("http://localhost:8983/solr");
SolrQuery query = new SolrQuery(":");
QueryRequest req = new QueryRequest(query);
NamedList<Object> resp = server.request(req);
String responseString = resp.get("response");