Don't feel bad about using something like 267 DPI instead of 300 DPI. You will not know the difference. Here's a tip out of Photoshop for photographers. If you need to enlarge a photo. Upsize in increments of 110% at a time. Not quite sure why, the book says so but I have done it several times and it works wonderfully.

I'd also suggest using Bicubic Smoother option when you do. PS resampling has been much improved, and many say it works as well as Genuine Fractals, when you resize in increments as suggested.

George makes a good point. Using Bicubic Smoother when enlarging will give you nearly identical results to most of the enlarging programs and to the 10% method.
If you're printing to an Epson, I would go for either 300 or 240 dpi for best results.
I have an article in the Mar/Apr eDigital Photo magazine on resizing in general - http://www.edigitalphoto.com/tips_techniques/0305printing/

If you'll be resizing after conversion, I suggest converting to TIFF. With JPEG you're already throwing out data on the initial save and when upsizing you'll want the start with the most data possible.

A couple of suggestions. Do most of your upsizing in the RAW conversion it works better. Second, to upsize a psd or tiff, repeat (via "actions") a smaller percentage (3-5%) size increase using BiCubic Smoother. Thirdly, I use 180ppi as my minimum resolution for Epson printers. Their RIP (drivers) make anything higher unnoticeable. I have never seen a difference with higher resolutions and have compared them to prints with double that (360ppi). This has been confirmed by many others as well for high end photographic printing with everything from the 2000P and the 9500 to the 4000 and the 9600. Does anyone out there have evidence to the contrary?
NR

If you use the Canon Raw converter or Photoshop converter it will not make too much of a difference. What you are looking at is pure pixel to dpi balance. The 10% Rule and using bicubic is all sound advice however I would recommend using it as a last resort if possible.

I have a 20D, your pixels will be 3504 x 2336 at the highest quality jpeg or RAW format. This will not change or SHOULD NOT change when doing any of your conversions or resizing if you want to maintain as close to lossless resolution with minimum defects/artifacts being introduced.

Now it all comes down to how you resize within the "Image Size" option of Photoshop.

In the Image resize you have two sections of information.

Pixel dimensions and Document size.

Both width and height information will directly effect each other. If you change the width and height in pixel dimensions it will change the document size and vice versus.

If you change resolution under document size you will see that it affects only the pixel dimensions and not document size.

Remember a pixel is a pixel. If you add pixels, making the pixel dimensions higher than 3504 x 2336, you take the chance of introducing artifacts. Again within 10% it will be minimal however see if you can make changes without having to do so if your print size needs will allow for it.

You are printing for an 11 x 16 print size. In an attempt to achieve the best your camera can produce change the resolution to 200dpi and then change your pixel dimensions back to 3504 x 2336. You will now notice that your document size goes to 17.51 x 11.673 inches. Which is very close to 11 x 16. My current explanation of this process is using a horizontal image.

Now change your dpi to 210 and make sure you change back the pixel dimensions to 3504 x 2336 you will now see your document size change to 16.676 x 11.117 inches. This will be your target. So for optimum results this is the setting you want to use to produce your 11 x 16 inch prints with as little loss technically. But if you want to increase the resolution and in turn the pixel count by 10%, you may get minimal artifacting but your print will not really appear much better. So why take the chance and waste the ink?

Some may agree with me

Regarding the resizing in 10% increments, a quote from "How to Wow: Photoshop for Photography" by Jac Davis and Ben Willmore:

"You may have read about a technique for enlarging images that involves scaling the image up in small increments to achieve high quality. That technique was invented before the Bicubic Smoother feature was available and for most situations, is no longer necessary."

So if you are using CS (or maybe CS2 at this point!) you may want to skip the small steps and use the Bicubic Smoother option.

Radar

Quote:
Regarding the resizing in 10% increments, a quote from "How to Wow: Photoshop for Photography" by Jac Davis and Ben Willmore:

"You may have read about a technique for enlarging images that involves scaling the image up in small increments to achieve high quality. That technique was invented before the Bicubic Smoother feature was available and for most situations, is no longer necessary."

So if you are using CS (or maybe CS2 at this point!) you may want to skip the small steps and use the Bicubic Smoother option.

Radar

Agreed. With the Bicubic smoother option, the 110% method is not needed. I shoot RAW and in ACR set the dpi for 300. Then re-size the converted ( now tiff ) file to the desired print size. I use bicubic smoother if the output is larger, bicubic sharper for down-sizing. Last step prior to printing is sharpening. I print to an Epson 2200.