I'm with N4 on this one.
Any time I've ever added a synthesized layer just recording the line-in signal my ears end up tripping out. It makes a lot of (scientific) sense though.
Think about the spatial properties of what ever sample you want to use. You can HEAR the room, you can hear things that are natural. The imperfections of the physical space what ever musicians were in when they recorded. When you record line-in, you miss out on that completely (and then try to make up for that by using software reverb).
THIS is where things hardly gel.
Listen to the intro "When the Levee Breaks". Everyone knows the song and there is a TON of ROOM in the intro drums. Sample that - now - take your synth line-in and jam with it. FEEL how WRONG that sounds? Hear how WRONG that FEELS? That's called "psychoacoustics".
http://en.wikipedia.org/wiki/Psychoacoustics If you mic'd your synthesizer and played a riff along with those drums, it may sound less wrong. It's all about the natural balance of things. (Un)Fortunately, due to survival instincts, we'll never be able to truly trick our ears/brains into letting something like ROOMY drums and SUFFOCATED synth lines to ever sound really "right". And I don't know if I would ever want it any other way.
Hope this helped someone!
Peace!