I'll have to think about this one a bit, especially after I get home to my books.
But on first glance it seems to me that the alkalinity of the first bath is high enough the a lot of development will take place. The sulfite isn't enough to be strongly solvent (that's usually said to need 100 g/liter), but it still looks a bit like D-23 with glycin.
The alkalinity is more dependent on the type of alkali than the amount (within reasonable limits), so there is no significant difference in the development taking place with 35g or 100g - or even 10g - of sulfite.
Fx-2 is an interesting choise for "splitting". I have been thinking along similar paths, but (still without a single test) am now leaning towards a metol-pyrocat developer. I'm also experimenting with replacing metaborate with bicarbonate for availability reasons (bicarbonate: supermarket, metaborate: mixed from borax - chem. supply house - and lye - supermarket).