Refer to http://futureplanets.blogspot.com/2009/07/venus-flagship-proposal.html for a description of the combined Flagship proposal.
NASA's Venus Science and Technology Definition Team officially recommends a mission costing (by their minimum estimate) $2.7 billion -- so expensive that Congress almost certainly won't back it. However, the STDT is quite explicit in saying that this plan is designed to be flexible:
"It is evident...that the STDT-recommended multi-element mission architecture has the highest science Figure of Merit [compared to other proposed combinations of missions, once one rules out truly costly missions involving the development of high-temperature electronics and/or refrigeration systems] and provides flexibility for payload accommodation on the various mission architecture elements. This allows for scalability in response to mission cost cap changes, and readily lends itself to international collaboration because partners can take responsibility for different elements that are highly independent." (pg. 283)
The obvious question is: where should the fracture line be set if NASA does indeed decide to advocate cheaper separate Venus missions? The most obvious answer is to split off the Venusian landers from the Venus orbiter and balloons. The balloons, after all, will provide far more science data if they're accompanied by a com relay orbiter than if they have to rely entirely on Direct-to-Earth communications during their month-long mission -- whereas the landers, which need to maintain a radio link for only 6 hours apiece (including their one-hour descent to the ground), could rely for their com relay on their carrier spacecraft during the hours of its nonstop flyby of Venus. Indeed, the only scientific argument for sending the balloons and the landers to Venus together (since they carry similar atmospheric instrumentation) is that it would allow simultaneous atmospheric measurements at the near-surface and at 55 km altitude several hundred km away as the balloons are blown away horizontally by Venus' high-speed superrotation winds at that altitude. This is a pretty weak synergy.
The interlinking of Orbiter with Balloons makes far more sense -- not just because of the vast increase in balloon data that it would make possible, but because of the Orbiter's ability to make instantaneous observations of Venus' meteorology and cloud patterns in a very wide area centered around each balloon's current location. The STDT study concludes that the cost of such a mission -- combined with two entry probes that would make atmospheric observations all the way to the surface but not survive their landings -- would cost $1.3 billion, only half the cost of the total Flagship mission. This is still above the cost cap for a New Frontiers mission -- but not far above it, if you ditch the entry probes. Another possible cost-cutting measure would be to eliminate the heavy and power-consuming SAR from the Orbiter in this mission and instead fly that on a separate orbiter mission.
Kevin Baines' "VALOR Plus" concept ( http://smartech.gatech.edu/
As for flying the landers by themselves on a flyby probe carrier with com relay: Larry Esposito's "SAGE" proposal for the first New Frontiers announcement of opportunity ( http://trs-new.jpl.nasa.gov/
Regarding Van's notes on the main reason why the STDT Landers are more expensive than those in "SAGE" -- the longer lifetime of the STDT Landers because they would drill up and analyze two Venusian surface samples -- descoping Venus landers to that shorter surface lifetime would bite into their scientific return to some degree, but certainly not catastrophically.
In short: the proposed Venus STDT mission -- while almost certainly much too expensive to be flown as a single mission -- can probably be broken up pretty easily into three separate missions (Balloons, Landers, and SAR Orbiter) which would produce almost as much science return as the original unified mission, and which would, if thus separated, probably be individually flyable within the New Frontiers cost cap at completely separate launch windows. And the very detailed and useful STDT report can still be used to provide science background and engineering design work for such a fragmented mission.