I'm only including VBM that's changed since the last update. There were no LA county ballots, so nothing changed there and many other congressional districts remained the same.
CA-3: Drops from D+6 to D+5, roughly the same as 2012. The first ballots were D+12. John Garamendi won by 8% that year, but this is a small enough advantage that this race should at least be on the radar.
CA-7: Drops from D+1 to even. It was D+2 in 2012 when Ami Bera won by 3%. This one will be tight.
CA-36: Back up to R+2 instead of R+1. If it moves a little more Republican Brian Nestande is in this thing. While UT San Diego has commissioned multiple polls for CA-52, no one has polled this district.
CA-47: The VBM returns went from R+2 to R+3, but that's a little cheating. They only added Orange County VBMs, which is the Republican part of the district. Since Andrew Whallon has spent $52k the entire cycle it's hard to take his candidacy seriously.
SD-12: Moved from D+7 to D+6.
SD-14: The VBM return is now D+19 down from a high of D+24.
SD-34: Moves back to R+9 from R+8.
AD-21: VBM returns are now D+5 up from D+3, good news for incumbent Adam Gray.
AD-36: VBMs are now R+11, after being only R+6 a few days ago. That makes it tougher for incumbent Steve Fox.
AD-40: Returns are now R+7, after being R+4 a few days ago. This is good news for Republican Marc Steinorth.
AD-65: VBM returns are now R+11. If this differential continues incumbent Sharon Quirk-Silva is doomed.
A big thank you to PDI for compiling this data.