A) Leviathan. At lower levels of play, Toshi sets up that extra level of mind game that opens things up for the already-chaotic PE games.
B) Exploratory Romp. Starlight is playable, in the proper context, and its cost can be neutralized now with Pancakes.
C) Test Ground. Nobody looked at this and said anything more positive than "Huh?"
D) Salvage. 0 Strength, low trace strength, Code Gate (therefore Yog-able). Even before you look at the challenges of making it function to do anything, it already sucks.
E) NeoTokyo Grid. As the game gets more silver-bullet-dependent, the ability to put the irrelevant silver bullets into the opponents hand could become more useful.
F) Disrupter. Data Hound is bad, but could be impactful if everything fell exactly into place.
G) Hard at Work. If it was 3 credits, it might be good, but 2 for the loss of flexibility is pretty mediocre. And as little good as Zed is, it could be huge if you could surprise the runner at exactly the right time.
H) Helium-3 Deposit. Corp cards with power counters you want are rare. Sure, if you can land a Data Raven, Mamba, or Viktor 2.0, it'd be good, but that's pretty narrow. So is Hellion, but still....