Giving up on prospects is a dangerous business. Just when you think a guy sucks, he'll break out and surprise. It happens again and again. Remember the LenDale White hate this offseason? It was pretty similar to the Larry Johnson hate a few years back, which was pretty similar to the Thomas Jones hate a few years before that. Now all the haters are :

because those guys have proven (or in White's case "are proving") that they can play ball.
People like instant results. When their rookies are slow to develop, owners will often lose patience and sell at a loss. This is a mistake. If all you did was judge a guy based on his first 1-3 seasons, you'd miss out on a huge chunk of the league's future stars. Chad Johnson was a nobody after his rookie season. Santana Moss was a flop. Plaxico Burress looked absolutely terrible. The list goes on and on.
The thing people sometimes fail to realize is that different players develop at different rates. So while Jacoby Jones might be making a smoother transition to the NFL than Ted Ginn, that doesn't mean he's a better player. Maybe Ginn just hasn't had his chance yet. Maybe he's learning the ropes. Maybe he'll be a top 15 WR in two years.
My personal philosophy is to try to keep my highly-drafted prospects on my team until I'm pretty sure they suck. Sometimes it's obvious from a player's first game that he's going to be a superstar (Randy Moss). But usually, it takes 2-4 years before you really know what you have in a player. If you're not willing to allow him those years to develop, why did you draft him in the first place? I think you owe it to your team to give your prospects a chance to bloom.
Obviously stockpiling prospects can cause a roster squeeze, but there are plenty of ways to cope with this problem. One thing that I like to do is convert my assets into draft picks. This past offseason I had a bit of a roster overload in the Hyperactive league, so I traded Matt Jones for a 2008 1st round pick and Wali Lundy for a 2008 3rd round pick. This allowed me to retain their value (in the form of draft picks) while also freeing up two roster spots that I ultimately filled with sleepers. Simple mathematics shows that this is a profitable maneuver.
Let A = value of Matt Jones = value of 1st round rookie pick (because we're assuming that it's a fair trade i.e. the value of the pick I receive equals the value of the player I give)
Let B = value of Wali Lundy = value of 3rd round rookie pick (because we're assuming that it's a fair trade)
Let C = value of the best player on the waiver wire
If I keep Jones and Lundy, I have A + B.
If I trade Jones and Lundy for fair market value, I have A + B. But since I also have two open roster spots to use on waiver moves, I actually have more than A + B. I have A + B + C + C.
As you can see, the total value of the second scenario is greater than the total value of the first. In plain English, by converting your players into draft picks you retain their value while also allowing your team to acquire additional value through waiver wire pick-ups.
So if your roster is loaded and you don't want to drop any of your developmental "project" type players, you can usually find a way to make a trade that allows you to cut some weight without losing value. Consider your roster as consisting of three types of players:
Group A: Starters. You need these players. They are important to your immediate success. You do not want to lose them.
Group B: Prospects. Prospects who are too valuable to cut. Guys like Sidney Rice and Robert Meachem fit this mold.
Group C: Flyers. You like these guys, but they don't have any trade value.
Now let's pretend the roster cutdown date is approaching. Since you probably want to keep your group C guys to see if they pan out and since they don't have any trade value, the proper move here is usually to convert your group B guys into future draft picks. This allows you to keep their value on your roster while also allowing you to keep your group C players.
Now some people might say, "Well aren't you selling low on Robert Meachem if you trade him before he develops?" Not really. You're selling him at market value. And the pick you get in return has value. This isn't a "win now" or "win later" move. It's a "win whenever" move because whenever you want to, you can convert your draft picks right back into player currency.
So if I trade all of my depth for future draft picks, I'm not really hurting my depth. That's because the picks I acquired have value. I can flip them for veteran depth at a moment's notice if I feel like it.
So in conclusion, I recommend keeping your developmental prospects on your roster until you have a good idea of their true value. If you absolutely must make roster space, it's better to trade one of your middle tier prospects for a pick than it is to drop one of your lesser prospects outright without getting anything in return.