Hi Ateeth,
Glad to hear from you. Hope you had a good time at GEC20.
For the Flow Statistics part we actually used the extended version of the topology you described above with 2 switches and 6 hosts. However, the Flow Statistics was only collected from one of the switches as an example. The same application script can be used for the scenario you describe as well.
Here is a link to the tutorial where we presented the Flow Statistics measurement.
To answer the next part of your question, you can get any information that your controller gives you into LabWiki. For the application described in the link above, we get the throughput on each interface of the switch and group them by ip addresses using the tag "flow_ip.nw_src". Similarly any other field of the message packet maybe used instead. The other fields you can use (incl. MAC address, dpid and so on) are shown here:
For this example, we used the Trema controller APIs (http://trema.github.io/trema/). You could potentially use any other APIs you prefer, get the message contents of your flow packet, write these to a CSV-formatted file and use LabWiki to visualize your experiment.
Hope this helps. Please let me know if you have any more questions.