From a router firmware programmer... NO!!!!
First, DD is not more stable than the stock firmware. In most instances it's actually less stable, unless you get one of the paid versions. The QA that goes into the free builds is minimal at best. I've seen this myself on every single router (and there have been many) that I've tested on with the exception of the original reason for DD's development. That's the Linksys WRT54G <v3 and 54GL. The problem is, people have equated its performance on that model to its performance on anything that supports it. Lemme tell ya, it ain't. Being told otherwise is simply a result of talking to people who didn't have their stuff set up correctly to begin with. Some will argue that companies like Buffalo are using DD stock for a reason, but they neglect the fact that Buffalo's versions are paid for and need to have a certain degree of quality assurance.
Second, DD's overclocking functions are total BS. The biggest effect realized by an increase in the input power to the RF radios is shorter life and higher operating temps. The single best way to increase effective range and bandwidth is with longer antennas and/or using antennas of more optimized forms for your particular needs. A better way to get higher performance is to increase the input from the transformer, not to increase the current draw of the radios. That would provide a more stable electrical signal, which helps the consistency of the wireless transmission.
My very last experience with DD was on a network of 3 WRT350N routers. For single access point usage, the stock Linksys firmware was solid. Never a problem, never a hiccup. With DD, even following all the BS you read on their forums about doing 30-30-30 resets and saving in specific orders (obviously these people don't understand NVRAM or they'd know the order is meaningless), it was just not stable on any of their recommended builds. The situation was even worse when the network was run with a single AP and two repeater bridges. As such I'm convinced that their implementation of repeater bridge mode is buggy.
In the end, I found covering a huge 3 family house with a single Netgear WNDR3800 to be more stable, higher performing, and further covering solution than jerking around with DD-enabled repeaters. Again, back in the day of the 54G, DD and Tomato and OpenWRT were great. It gave us features and performance that wasn't possible on stock. Those days are over. Unless you absolutely want a VPN concentrator or just want to increase the capabilities of an old router, I see no reason to use DD on basically any recent router model.
Lastly, real advanced users that like to tweak their stuff and own a DD-supported model don't need DD. Supporting DD simply means that the router is using a Linux kernel and has the requisite amount of RAM. Clicking some GUIs isn't advanced, writing your own firmware is. It's more for getting features that you want but can't code yourself.
That's my rant and I'm sticking to it...
