1) if the GN connection to Thameslink isn't used, it won't be because of the direct cost of installing it, but rather because the London Bridge/Borough work gets scaled down, cancelled or delayed. Without the new flyovers etc south of the river, Thameslink only has enough capacity for the current MML commuter service.
2) assuming TL to GN does go ahead, the trains on it *will* be semifasts, because the Thameslink stock will be 100mph and Network Rail and DfT want to remove all sub-125mph trains from the fast lines.
3) splitting the Hertford service from the Northern City Line service would be crazy, as the vast majority of demand on that route is for commuter services to the City. The Underground from Finsbury Park is already overloaded, and reversing Northern City trains at Finsbury Park would either massively eat platform capacity (if a cross-platform interchange were provided with KX-bound trains), or require massive use of the overcrowded, narrow subway. And the link from FP to the NLL is only single track with no flyover, so would be operationally nightmarish-to-impossible to use.
4) All the relevant franchises (SN, SE and FCC) will be re-let and re-organised around the time Thameslink opens. It's not yet clear what the new map will be - the operator that runs Thameslink services may or may not also run GN Moorgate services, SN London Bridge services, SE London Bridge services, etc.
At that point, Thameslink won't use 319s but rather newly developed stock (because current stock doesn't have the accelleration, braking and ATO capabilities to meet the proposed timetable between Blackfriars and St Pancras). The Sutton loop won't need those capabilities, so it's most likely to end up operated by some kind of suburban Electrostar.