1) The secondary response requires CD4+ T cells to activate memory B cells. That first paper actually gives some evidence that some of the rapidity could arise because T cells and memory B cells are in very close proximity to each other in germinal centers.
2) Yes. The affinity of antibodies increases during the _initial_ infection, both through isotype switching and affinity maturation, producing far superior antibodies. Upon reactivation, memory B cells can undergo further somatic hupermutation.