That said, for my side project ( http://artfulmac.com ) it is absolutely worth it to me as a solo developer to pay Apple their 30% to have them handle fulfillment (including system-integrated update notifications), payment processing (with currency conversion), tax form creation, and refunds. It's a trade of financial overhead for time. The App Store also conveys a sense of trust to consumers.
I think one aspect that gets easily lost in the debate is the long-term impact of having your app on the MAS (exclusively or not). You need to consider the value of being able to provide trials (i.e., do not force potential customers to spend money upfront on software that might not fulfil their requirements), discounts for your long term repeat customers, the ability to communicate directly with them and resolve their problems.
If you want to run a sustainable business, whether the MAS is a net positive becomes a question mark. I have not seen much supporting data to suggest that the MAS is a net positive if we compare exclusive direct vs exclusive MAS over a long period of time.
Anecdotally, from the data I have on hand (conversations with customers and friends), almost no one discovered their Mac apps from the MAS (there's some bias in my sample, of course). If they needed the software, they would have bought it - whether it's outside or not doesn't matter. If you need a tool for your job (the type of software I'm interesting in making), you're not going pass on it just cause you have to enter your credit card. For other types of software, the more disposable ones, convenience plays a much bigger role.
And most importantly of all, it will all depend on the particular type of software that you're selling. In my experience, the MAS works great for type of software that's mostly consumable - cheap, disposable, no long term plan to keep it updated. For other types, where you plan to run a sustainable business over many years - I'm not convinced.
While you'll probably do less units selling directly, you'll be getting a significantly higher proportion of the sale price. It's important to consider all aspects, not just raw revenue over a short period of time and make a decision from there - for some apps, the MAS would be the right way, for others - not so much. Now, deciding which way to go is a much, much harder question and I'm afraid there's no easy way to come up with a definitive answer.
Their handling of iOS and Mac apps seems indifferent at best and downright incompetent at best. In any case I've left their proprietary ghetto and gone back to coding for the web.
There's also the fact being on the App Store is a form of advertising. I think this is more obvious in Steam, which basically creates markets for games that had none before (usually because indie game developers are not necessarily good marketers). I know I've "browsed" the App store to look at things.
In these sales, it's not getting only 70% of revenue you would have otherwise, it's getting money you would not have gotten at all if you weren't present on the store.
Putting your software on the App Store or Steam or similar (notably not Google Play) means you're just getting paid royalties, not selling a product to individual customers.
"You are the merchant of record for Products you sell through the Store." I'd bet that 95% of smaller Android app developers have not hired an accountant to review every last detail, and therefore have failed to pay every cent of tax they owe. Unless you're very serious about starting a real business, it's a total nightmare.
Furthermore, if you're UK based and use a non-UK marketplace / reseller, then the revenue you receive from the reseller lies outside the scope of UK VAT and you don't have to do anything 
The reason why most EU developers on the App Stores don't deal with VAT is because the vendor of record is Apple Luxembourg, thus lies outside the scope. And that's precisely why Apple requires VAT invoices from Luxembourg-based developers.
Important issues with sandboxing were never resolved. Meanwhile existing APIs keep being moved to the restricted list, so as a developer you can't even count on your product being able to stay on the Mac App Store.
While it's not the end of the world, it's just plain annoying having to remove such a common sense feature. The consequence of this means that if you were to distribute a web browser via the MAS, you will not be able to ask the user to set it as the default web browser. That's just a subpar experience.
My hypothesis for the added restriction is that it might be considered a security hole for an app to just set itself as a handler, for say, http links.
The correct way to have handled this particular situation was to introduce a new entitlement and define what content types an app might want to handle, which will get reviewed by the MAS review team. Unfortunately, the solution  is: "There is no solution; it can't be done anymore in the sandbox."
Berners-Lee's original web browser was developed on the NeXT (direct ancestor of OS X)... But if he wanted to take that 1991 browser and distribute it on the Mac App Store today, he would have to remove important functionality.
Which begs my original question: what good is a "Mac app store" that can't accommodate Mac apps?
Apple isn't alone in this, of course. The Windows 8 app store (Marketplace, or whatever it's called these days) is just as limited... But really, if the most apt comparison for the MAS is the Windows 8 store, that further illustrates just how badly Apple has blown it.
Target market? Well lets try the average user who is much better off with buying from Apple who already has their credit card than going through some little website and their pain in the ass credit card form, registration and what not.
That's not the greatest endorsement ever.
"Of course I do interesting stuff on my computer, but the average person doesn't. All they need is X."
Importantly, X is never defined by asking actual users, but is simply a distillation of prejudice about the needs of "the common people".
You keep using that word. I do not think it means what you think it means.
From Wikipedia: "A ghetto is a part of a city in which members of a minority group live, especially because of social, legal, or economic pressure." I don't see how that applies to the app store.
Maybe you're trying to use the slang definition of ghetto, in which case you want an adjective and not a noun. Try "The Mac App Store is ghetto, yo." But "The Mac App Store is a ghetto." I don't know what you're trying to say with that phrase.
Probably the best way to look at it would be to treat the Mac App Store as a secondary channel. If it was a secondary sales channel, then it's additional revenue and new customers which is a nice to have.
You are probably best off creating your own customer list and selling directly to customers over time away from the Mac App store. Or, you avoid it alltogether and figure out how to sell direct. The only downside there is you have to do your own advertising and use something like Gumroad or FastSpring or some homegrown Stripe solution for delivery and such.
Nothing is perfect and there are tradeoffs to both approaches.
"Apple has taken $600,000 (USD) of that in fees. Ouch!" is much more concrete than just "30% of revenue."
Well, there are providing a service in return for those fees as well as high-visibility amongst potential customers. Striking out on your own means a lot of work in getting people to your website.
# A message from your local software retailer who's doing really well by selling directly.
It would be nice if Apple would fix the search features on the Mac and iOS Apps stores so people can actually find my software.
11.13 Apps can read or play approved content (magazines, newspapers, books, audio, music, video) that is sold outside of the app, for which Apple will not receive any portion of the revenues, provided that the same content is also offered in the app using IAP at the same price or less than it is offered outside the app. This applies to both purchased content and subscriptions.
I think it is just too big risk to download from a small business which I don't know and run their application outside the sandbox.
So, I'd recommend that you have some version in MAS, otherwise you may lose customers.
I talked with local audio gear seller and he told me that we will start to see less audio products compatible with Apple as they just want to grab 20% of revenue.
If someone recommends an app store product, they'll link to apple's website.
If I sell products from my own site, and someone recommends it then they'll link to my site.
In general, inbound links are worth a lot more than individual product sales.