What are your criteria for quality when you first read through an adventure?
I look at two broad categories. Organization and content.
Organization. This mostly comes down to whether the designer understands that the module is a game accessory and treats it as such. Table of Contents and the index (or lack thereof) are the first things I look at. Is the ToC thorough or anemic? Is there an index? Does it look thorough or anemic?
Layout and design is the next thing I look at. Are the pages well done, is information easily findable, is everything you need to run an encounter there on one spread? The more info you have to flip pages to find the worse I’ll rate a module. Are the relevant maps on the same spread as the related encounter? If there’s a bigger map is there a quick way to tell how this map relates to the whole?
Summaries, check lists, and bullet points. If the writers use a lot of these, the module will most likely be better than average in the usability department. A summary of every scene is great; a summary of every important NPC is necessary. A summary of what's happening in the module is required; a long-winded explanation of the entire history of events leading up to the situation in the module is cool, but largely useless. It's especially maddening if this is the only summary given as you have to read it multiple times highlighting things and taking notes for use at the table.
In general, D&D 4E and OSR modules utterly dominate in this area. The designers treated the module like what it is: a game aid for the referee to use at the table to run for players at the table. Treating the module like a novel is a terrible mistake and leads to terrible organization.
Content. As for the module proper, the contents, that all depends on how much the module constrains player choice, assumes what they will do, how detailed the villain's plot is, and how modular the content is.
Player choice. If the module assumes they will turn left at a T intersection and places no info at all for them turning right, then it’s bad. It's not hard to place hooks that will pull players in certain directions but some modules don't even bother with that. Like the 5E Dragonlance module. After the first scene or two it's just assumed you will travel to this one place. No hook, no reason...that's just where the next leg of the adventure is...so that must be where you go. There's the whole world of Krynn you could theoretically choose from, but somehow you have to go there. If hooks are done badly it amounts to fetch quests for plot coupons, but even that is an improvement over...just randomly pick the right location to go to in the whole world. Is every step of the module scripted? That's incredibly bad design. You can usually spot a lack of player choice when you see lines like, "In this chapter the player characters will..." Nope. No prep survives contact with the players. That goes doubly so for prepped modules.
The villain's plot. The referee in an RPG controls the NPCs and the whole wide world. Not the PCs. So whatever story a module presents should focus on what the referee controls at the table, the NPCs and the world. What is the villain's plan? What will they do if/when that plan is disrupted? And how will the world react as the villain nears completion of their plan? What are the weak points of their plan? How can the PCs disrupt those plans? How will the villain react? Etc. Is none of this detailed? That's bad. The amount of this present in the module usually has a direct correlation to the above criteria, player choice. The more choice the players have, the more detailed the villain's plot needs to be. The less choice the players have, the less detailed the villain's plot needs to be.
Modular. Because most modules are linear with little player agency or details on the villain's plot, I tend to not run them. But I do strip them for parts. So the more modular individual scenes, NPCs, locations, magic items, etc are the better.