More and more, I think the approach of "add more artifact destruction to the board" is a futile way for most blue decks to fight Shops. Especially when you don't have access to Red for Ingot Chewer. For a long time, I was playing Talrand Gush. It was a completely solid deck, with a stable mana base and 8 or more legitimate sideboard cards for Shops. Still, it had an extremely low win percentage against Shops on the draw, and I would often lose games even when starting with a hand of like land, land, mox, Nature's Claim, FOW, etc.
None of this is breaking news, it's well established that a lot of decks have a really hard time winning against shops on the draw (even postboard). It doesn't matter if you have an infinite amount of 1 mana answers in your hand if you don't have the mana to play them (or worse, if Chalice is set at 1). That being said, what was becoming increasingly annoying to me were the situations where I did in fact have the mana to play my removal spells, and still lost. In many situations, you are barely surviving and scraping by to piece together a board that can finally Bolt a Lodestone, or Ancient Grudge a Sphere, or Hurk's to get rid of Chalice on 0 , and so on. But then another threat comes down and you are floundering to survive again, or perhaps finally get locked completely out of the game.
So when you have a small window to actually cast/resolve a spell against Workshops, it seems that a lot of decks would benefit a lot more from playing a card that will simply win the game instead of grind out incremental advantages. Of course, I'm referring to Oath of Druids. The recent success of the Burning Wish decks with Griselbrand attest to this tactic being viable, and it has been shown to work through the sideboard through some recent events. One such example is a Gush deck that won the LCV (incidentally, the second place finisher from this tournament also had an Oath sideboard):
Creatures [1]
1 Talrand, Sky Summoner
Instants [28]
1 Ancestral Recall
1 Brainstorm
1 Gifts Ungiven
1 Misdirection
1 Mystical Tutor
1 Noxious Revival
1 Rebuild
1 Vampiric Tutor
2 Brain Freeze
2 Flusterstorm
2 Remand
3 Mana Drain
3 Repeal
4 Force of Will
4 Gush
Sorceries [7]
1 Demonic Tutor
1 Merchant Scroll
1 Ponder
1 Regrowth
1 Time Walk
1 Timetwister
1 Yawgmoth's Will
Enchantments [1]
1 Fastbond
Artifacts [8]
1 Black Lotus
1 Lotus Petal
1 Mana Crypt
1 Mox Emerald
1 Mox Jet
1 Mox Sapphire
1 Nihil Spellbomb
1 Sol Ring
Lands [15]
1 Flooded Strand
1 Library of Alexandria
1 Polluted Delta
2 Misty Rainforest
2 Scalding Tarn
2 Tropical Island
3 Island
3 Underground Sea
Sideboard:
4 Forbidden Orchard
2 Rune-Scarred Demon
1 Tidespout Tyrant
2 Abrupt Decay
1 Hurkyl's Recall
4 Oath of Druids
1 Ravenous Trap
Utilizing this tactic only has a few requirements. Mainly, you should probably have access to Green mana outside of just Forbidden Orchard, and you need to not be reliant on playing creatures of your own. It would obviously be pretty stupid to have a Bob based deck and then bring in Oaths. Also, it would probably be more beneficial anyway for Bob decks (or decks with Wasteland, or Deathrite Shaman, etc) to stick with efficient removal spells since they are more apt at fighting Shops on the board than something like a Gush deck. Aside from those requirements, you also want to utilize this in a deck that already has an overwhelming win percentage against many (most?) of the other archetypes. The Oath plan eats up a ton of sideboard space, so ideally your maindeck would already be very good at beating blue, or have a very fast/consistent combo to outrace Dredge.
One specific application of an Oath board that I would like to bring up is in the Doomsday archetype. Doomsday is notably great against most blue builds, and fast enough to combo out against Dredge without needing to disrupt it. Furthermore, you can save a lot of space in your board by not running any of the Oath creatures, since you already have a two turn win (that can mostly ignore Grafdigger's Cage) with Oath + Laboratory Maniac. Here's what I've been running lately with some promising results: