Should we increase the node count for A2A testing of the Large ref arch?

[billclaytor]

This issue arose due to the perf-agent-group not being classified for Large. We should wait for results with classification to decide this. The actual transaction times are not expected to be different between Standard and Large; the results should confirm this.