There does seem to have been a hardware improvement to reduce Vampire drain between my P85, Sig SN:00037 and my P85D built last 2014. I left them both connected to power while I went on a 1.5 week trip with energy saving on and always connected checked. The P85 Sig was running 6.2-101.36.2, and the P85D was running 6.2-2.4.153. I checked on them a couple of times in the 10 days, but did have charge alerts on so I could see when charging happened.
Both cars started charging when they needed 9 rated miles and replaced the 9 rated miles. The P85 Sig charged 9 rated miles every 1.68 days or 5.38 rated miles per day, and the P85D charged 9 rated miles every 2.43 days or 3.70 rated miles per day. Even with the P85D rated miles being 5% more than the P85 Sig rated miles the P85D has a much less thirsty Vampire.
Put another way, the P85 Sig Vampire drank 1.79 kWh per day, and the P85D Vampire drank 1.29 kWh per day with energy saving on and always connected checked. This is still a very thirsty Vampire, but at least Tesla seems to be slowly taming the Vampire's thirst.