The Princeton model is tighter and its highest peak is around 320. Silver's model has always had at least one of its top three peaks between 340 and 360, even when her swing state and national poll numbers were falling. Both models currently predict a close EV (between 290 and 310), but 340 to 360 is hardky impossible.
Plouffe's working for Uber now, but he's still privy to HRC's internal polling. Acknowledges they still need to turnout millenials, skim enough Johnson voters to cap Johnson/Stein combined @ 8%, and have issues in IA & OH. Despite all that, still believes it will end up a 4-6 point race with Hillary sweeping PA, VA, NH, CO, NV, FL, NC, OH, and IA. IA, OH, and NC seem overly ambitious to me, but Hillary's in IA today for an early voting event. Plouffe did say Obama's internal polling on NV was low on both turnout and results both times.
Haven't heard anyone from the Clinton campaign say it publicly, but they have to be pleasantly surprised by the White educated support nationally and especially in swing states and are quietly capitalizing on that. Certainly aren't going to concede Blacks and Latinos, but they have enough data and are nimble enough to adjust on the fly.
Conway's a good pollster, but she willingly went to work for a deeply flawed asshole who refuses to take advice and execute reasonable electoral strategies. KAC definitely knows the Trump base alone isn't deep enough to win in swing states, so they have to add signifcant amounts of suburbanites and skim urban Blacks and Latinos. No way KAC advised Trump to trash Nieto and the Black Pastor after highly stage managed photo ops. Same with taking the bait on Cuban and inviting Gennifer Flowers or trashing Machado after a night's sleep. KAC's not a campaign manager, but a spokesperson who has no control over Trump. Doesn't matter if Trump refuses to listen, doesn't understand the strategy, or refuses to prepare, it all ends up in a clusterfuck.