Some issues to investigate:
1) You'll need to know the forward voltage of the LED to really analyze if it is appropriate. Fortunately, it's just a a diode. Put your multi meter on the diode setting, connect it up to your LED and read off the forward voltage, it'll probably be 2.0 ish, but you never know. I just measured one random one out of my parts box at 1.6 volts. Now Vcc on the IO board is 3.3, and this parts box LED would drop 1.6v, leaving 1.7v across the resistor. Pulling out our old friend V=IR, we get I = V/R = 1.7/200 = 8.6 mA, which should be fine, but maybe not terribly bright. 20 mA is a common value for 'normal' 3 or 5mm LEDs. More current = brighter LED, up to the capability of the LED, or of the driving device to source or sink that current.
2) Speaking of brightness, LED brightness varies greatly, you can get 20 mcd from LED A, and 1200 mcd form LED B. Up to 300 ish are fairly common. The indicator LEDs on the IO board BOM are in the 200 - 300 range. The TOSlink LED however, is much brighter at 900 ish mcd. You are using the LED to light up an optical fiber in your TOSlink cable, if it isn't sufficiently bright, you wont get a good result. The ones in the no name kit probably are at the lower end of the spectrum. Maybe they work, maybe not.
3) And, speaking of spectrum... TOSlink is spec'd at 650nM, which is basically red. Whether a random red LED will work will depend on what specific wavelength that LED is, and how much wiggle room the receiver side has.
All this is academic though, a couple hundred ohm resistor is probably ok-ish for any random 3 or 5mm LED you might pull out of a no-name kit. You may not have as much luck for the TOSlink LED though, it really needs to be quite bright, and a cheap kit probably isn't full of super-bright LEDs.
Also, hardcore EEs will probably tear this all apart, but it's accurate enough for what we are doing.