While waiting (with bated breath?) for the answers, I'll look at E. After 1H-1S-2D-3D I chose 3NT. It's imps and we are vul. At matchpoints I would pass. i doubt that 5D is making. It might, but I doubt it. And 3NT is obviously iffy. But it's vul at imps so i bid game in something.
With the bots it went 1H-1S-2d-3D-4D-4H-Pass. It's a 5-2 fit but maybe not a crazy idea. Off 1 with the layout that I used, as is 5D. I am sure that I would pass 3D at mps but here game seemed worth a shot. 3NT makes as I placed the cards for the bots. I was focused on producing an auction that began 1H-1S-2D-3D but still I might have placed the cards differently.
It may well be that 5D makes more often than 3NT.
I am hoping for a good score for my friend the robot.
Still no scoring from BWS so maybe I say a bit more about E.
We could have
KQTxx
xx
KQxx
xx
A
AQ983
AT62
JT4
Assuming Ds behave we see 3 spade tricjk, a heart trick and 4 diamond tricks which gives 8 tricks. of course there will be a ruff. Spades have potential, so a reasonable plan would be to ruff a spade in hand and hope this establishes spades. But four spade tricks and a ruff, plus a heart and four diamonds, totals ten tricks so we would still need the heart finesse. This means, I think, that we should forget about establishing spades, assume the heart finesse works and see how that goes. Maybe that's the route, not sure.
All in all, game seems iffy. Probably, or perhaps, bidding game is justified by vul at imps, but risky. Definitely a pass of 3D at mps I think. Since 2D by S was passable, and 3D by N was passable, it seems 4D by S should now be passable so that's an argument for bidding it if we think pard will pass when his 3D call was a minmaum 3D call.