This will of course depend on how one defines "health" of the format, but assuming that this is somehow related to diversity of deck archetypes here's my thoughts.
The new method of reporting a limited number of collated decks with a minimum of 10 cards different between decks for leagues make this data unusable for assessing the health of the format. It will show an arbitrary diversity of decks that is in no way associated with the meta percentages represented by those given decks.
This leaves us with the excellent break down of the weekly challenges that we have for some idea of the meta breakdown for mtgo. Though I feel this may be skewed due to the popularity and convenience of leagues.
With so few reporting's of significant statistical value for the paper meta it will be difficult to assess whether it will be heavily varied from the online meta. Of course you'll always have more budgetary issues, etc. with paper so that will additionally need to be accounted for.
Just finished up a league, wen't 3-2. Here's what I ran into:
L Jeskai (saw a single mentor and tinker)
L PO Storm
W Unpowered Hatebears
W Big Blue
I was running Ravager Shops with Wires replacing thorns. Already jumped into my next one. This is pretty great!