Yeah, it would be nice...however, until a map gets a large amount of downloads and votes it will be useless due to different tastes.

The Starcraft 2 custom Mod/Map download system is an example of how wrong a popularity/rating system can go.
Without some sort of organisation, different categories, what have you, a pure popularity system ensures that only the popular mods stay popular.

Last submitted / updated is one way to spot the active / maintained mods but that doesn't necessarily mean that a finished and wrapped up mod has to be
worse than a maintained one.
I have some "old" scripts out (other games) that are "not being maintained" because they have a clearly defined function and perform it perfectly.
There is no need to change that.

Still -filtering by submission / update date is a good way to find out "what's new" when you're not stopping by every day.

I do think that a rating system is a good thing but it should be a bonus. Some mods are buggy or imbalancing, no doubt.
It's more important to be able to narrow down the list with a search or filter for keywords so the displayed list does not contain 371 highly rated mods that you are not looking for.

The game is just way too new to have a polished mod yet. How can you polish a mod before the game itself? I am working on a mod, rather a total conversion, that will change Elemental into the world of Avatar: The Last Airbender. But currently, after slaving away in 3ds max, I need to sit on my models and animations, awaiting to see what the new Elemental patches will bring.

I think the way they've split mods in the reactor makes it absolutely useless. It's useless because 95% of the mods released so far DO NOT FIT INTO EITHER CATEGORY.

First, Maps - completely understandable, the only real category.

Second - Tiles - wtf? Who's going to upload single tiles? Tiles aren't even end-user, so what is the point of using the reactor for that? There is NO point.

Quests - Admirable, but I have my doubts. Will people be interested in downloading single quests? Quests that, quite possibly, require you to download a separate tile as well? Quests also require a huge amount of bug-searching. How will quality be maintained?

Campaigns - effectively a scripted map, as far as I can tell. Is ANYONE working on a scripted map that ISN'T also part of a mod?

This is how I'd like to see it

Maps - obvious

Tidbits - smaller mods that add a couple of weapons, a new urxen skin, some bugfixes, new spell effects. Maybe quest packs.

Races - complete race mods, like undead rising (I think this category will be very popular, since the original races are so fucking boring and generic - yes, an undead race is LESS generic than what we have!)

Conversions - mods that completely change the game, such as Dragonlance, Airbender, LOTR.

Mechanics - python mods that alter, or add to, the way the game functions

THEN I'd like to see TAGS for each mod. Very basic tags, such as - "weapons", "spells", "skin", "quest". These should be clickable, and when clicked on should display ONLY mods with these tags.

As for what each mod should have; thumb up, thumb down, comments from users (that can also be thumbed). Last updated. Clickable link to webpage.

Edit: I've thought about it, and it seems the reactor is just meant to add content to the parts that the game fell short on (too few quests - too few maps - too few campaigns). Not accomodate mods at all.

I also understand that the categories listed by reactor are all "compatible" with each other - ie, you could download all of them and it wouldn't be a problem. That's a good thing, and also completely different from how "real" mods work.

That's why tags are so important. You'll never be able to describe mods, that can contain any variety of things, with categories. Tags can describe the contents without limiting it to a category.