We all had time to be razzled and dazzled by the good boys. You can have a legion of dogs at your command to destroy your foes and zoom about the map.
But these aren't even okay boys, let alone good boys.
As a 4 stacked ally, Sled dogs create what veterans of the TCG community know of as a 'Rube Goldberg Deck.' Your making a ton of effort and investing time and resources into a combo, which can be fine, but the payout isn't proportional at all to what you are getting. And understanding why they are bad is important to understanding what makes a combo deck actually good.
Sled dogs is taking up an ally slot, 12 resources, 4 cards, and 4 actions to come online. If your using the rod of animal magnetism that also comes out to a relic slot, and adds an action and card, but reduces the cost of the dogs by effectively 2, if you get the rod first. Because actions and cards are worth more than resources, this comes out to you 'behind' to play the rod in the best case scenario, and worst case if you need it to play your last two dogs because you found the first two before the rod, your looking at a grand total of 12 resources, 5 actions, and 5 cards to get set up for...
A damage 4 attack that can only be used once per turn, or the ability to move 2-4 spaces for one action, which doesn't even have the pathfinder or shortcut effect of actually saving you time to do things that turn (like fight) unless you 100% had to move all those actions anyway.
And its worse because this is a 5 card combo that really needs to be fully completed to get a payoff better than A worse Mauser. Would you pay your ally slot and 8 resources for an infinite ammo Mauser or a weird pathfinder pseudo-downgrade that only is usable if you move two spaces a turn? Probably not. But your going to be spending most of any given scenario doing exactly that. By the time your dogos are actually starting to resemble something better than a 3-4 cost level 0 weapon you could have just... played... your likely almost done.
Dogos show you what a combo element shouldn't be: Low payoff, high effort, and 'backend heavy' all at the same time. There are plenty of clunky combos in Arkham that are varying degrees of good, but in general for a combo to 'make sense' you want the payoff to be something that actively helps you win the game and thus worth time and deck inconsistency, relatively low effort (its ok that your going to go out of your way a BIT to do the combo, but not 20+ some odd cards, resources, and actions expensive), and should start paying off before it even is 'ready to go.'
It is ok if you don't have all of these qualities, but having all 3 is a bad sign. Sign Magick isn't amazing, it is a clunkier LeoDeLuca at first, but its offering you a bonus action every turn the second you get two relevant spells and it into play, which is probably just a 3 resources tax for ya, a card, and an action, and then it can scale up from there. So while its low payout, its also not a huge sacrifice (your playing spells anyway) and not too high in opportunity cost (Its half the price of leo, and while mystics have decent hand options they aren't going to lose sleep over it). Pendant of the Queen takes a lot time to do anything, but it is very cheap to do (your doing something you should be doing all the time anyway: Getting cards in seeker), and the reward is huge (extra testless actions and the ability to teleport anywhere in an action-less manner to boot). And while some recursive ally stuff in Guardian using things like Agency Backup and Beat Cop are going to take mountains of resources and a lot of your actions, they are going to be doing work from the second you initially play the card (neither of these cards require other combo elements to start to do something, and the payoff for the work is very proportional to it (who doesn't want effectively infinite free testless, action-less damage and cluevering?).
This is ultimately what makes dogos a 'Rube Goldberg' strategy: you are making a huge expensive machine that takes five minutes to set up and operate all to flick your lightswitch. Your paying a ton of money, slots, cards, and actions to... get a relatively janky weapon or movement too. It is a 'nothingburger' combo because what they give you doesn't really help you win the game (payoff), takes a ton of work and resources (effort) doesn't turn on until you get a lot of dogos (back-heavy). So, sadly, these dogos just don't make sense unless you know the scenario is going to demand you move 2+ spaces a turn constantly. At which point they kinda start to make sense, because while they don't help you 'stuff a turn full' like the aforementioned shortcut and pathfinder, if your always taking those actions even outside of reacting to threats you probably will save a lot of time in a scenario if you get two dogs and are using them most turns (and aren't able to slot a better action compression tool like Leo, Running Shoes, or those aforementioned cards) they will do... alright... but I think any campaign where sled dogs is a huge difference maker is going to be a deeply unfun campaign anyway.
And sure, they are neutral, but if you don't have a 'weapon' better than the sled dogs, you probably should be running other tools to defend yourself. Like I am sure Mandy COULD get the dogs out to serve as her 'offensive tool' but that isn't going to be a remotely good use of Mandy's time compared to the aforementioned Pendant, or the plethora of Seeker evasion tools that exist now.
EDIT: It turns out harshing folks mellow of leading a legion of dogos understandably ruffled some feathers. Some common counterpoints to the good dogos and why they don't make sense.
"You can't compare them to an XP card." I absolutely can, unless we are comparing the 2 dogs, which are the only ones who anyone but Leo can run. As a 0 XP card, dogs compare poorly to most weapons, as most weapons already grant you a +2 damage bonus. While dogs don't run out of ammo, they are once a turn, which is really bad for a 2 damage weapon, as anyone who has run a Mauser outside a 'success by X' test can tell you. The reason spammable, high/infinite 2 damage weapons are good is you can freely push up to 6 damage in a turn. Dogs can't do that, so not only is it action intensive to get the dog past a knife but you need to either have innate evade to handle 3+ health, or a secondary weapon, just pushing up the insane cost of dogs more. At 3-4, you can then directly compare them to XP cards (Both weapons, and 'assemble the pieces' cards like Segment of Onyx), and they REALLY don't do well there.
"What about the soak?": We all know highly action inefficient healing isn't good in Arkham, and the dogos are that. 2/2 is a very common ally statline on 'your not supposed to kill them' utility allies for a reason: it pushes the ally to being about as inefficient as something like Emergency Aid. Is emergency aid totally unusable? No, but you absolutely wouldn't rush to play it. Unless an ally is one shot or has unusually good soak values (ex: Peter, who effectively is infinite healing), the soak value is not the 'good' part of an ally. Would you run an ally with 2 health, 2 sanity, 3 resource cost, and no text box?
No one plays 2/2 allies as 'soak first' because the action economy is terrible. 2 Dogs takes 2 actions to play, 2 cards drawn (which are a little less than an action), and 6 resources (Valued at their lowest, which favors the dogs the most, a resource is going to be fractional of an action at around .75 resources to an action). This makes playing 2/2 allies that you need alive (And if you don't need the dogs alive play literally any ally that gives you either a come into play or on death benefit) at 3 cost worse tempo than first aid (Your looking at around .6 healing per-action spent from First Aid viewing resources as expensive as possible, with about .57 for dogs valuing resources as little as possible). First aid is probably the worst healing card in the game.
This means the 'text value' of the dog needs to be good enough not to just pass first aid, but to get past equivalent card effects. A lot of effects boost damage and to hit rates at a level similar to a fully assembled dog team at far less resources, decksearch effort, and action investment: Custom ammo, beat cop, enchant weapon, Delilah O'Rourke, Acidic Ichor, an upgraded shriveling, and the bows all will give the characters who can take them a larger effective boost than the dogs (outside of custom ammo obviously, but guardians have comically little need for the dogs), and only require one card and way less resource investment. You can (generally) evaluate a +1 damage for the rest of the game as about 3 resources and 1 card/action on top of your weapon, and some of these cards (such as the bows, or custom ammo) are doing even better than that. So, in reality, your dogs are just giving you 1 free 'first aid' in addition to the 2 you played first once you get to 3 dogs to match "weapon+enhancement." Would you play 2 first aids if, after clearing the first two out, you got a third one (Completely ignoring that you need to see 21 cards to even find the 3rd dog 50% of the time) finished for free? Almost certainly not.