Matrix_1970, Datastores have tons of useful information. I make several dashboards for datastores. One I always use is heat maps - I call the big four from my old integrien days. I put key metrics in each of four heat maps on a screen. Within each map I would label one something like latency, cpu, etc, then put several metrics in each view. Sometimes I will use the mode for instance and sometimes use the general mode to create a sized by and colored by combination. I do that for latency and it makes the block size large for long latency times which catches the eye. I also have my vmware team provide the array to luns to group those together as Applications and each tier is a group of devices that the vmware team tells me belongs together. Those are good for application displays which can be part of one of the datastore dashboards. Maybe add a metrics which is fed by application widget. This is interactions. The manual is good at helping understand how these work and which widgets can feed and which cannot. The v5 gettin started guide talks about dashboards.
Cannot help on q2. as I have similar questions. Perhaps vmware will put together an advanced vcops user class. E1 does great powerpoints that are extremely useful.