Advantage:
Any layer blending or styles will (for the most part) keep their RGB appearance; though you will notice some of the usual red and blue color shifts and sometimes you’ll lose a bit of intensity in the drop shadow/glows/bevel effects.
Disadvantage:
You lose the editability of your layers. If you convert to CMYK without flattening, you will likely notice substantial differences to any layer blending or styles.
Alternative to flattening:
Well, I have two:
#1: Unless you are heavily dependent upon filters that use the RGB color space, design in CMYK.
#2: Instead of flattening and converting from RGB to CMYK, I find that I retain more of the original color values if I copy merged from the RGB document, create a new CMYK doc and paste the merged RGB image into the CMYK file. I then save as TIFF.
Ideally, I would think that merging and flattening would yield the same results, but for whatever reason, I find that copy merged lets me retain better color.
#1: Unless you are heavily dependent upon filters that use the RGB color space, design in CMYK.
Working in RGB allows you to repurpose the image for any output from web to press. Working in CMYK sort of pins your output to one cmyk device.
