Ozzy,
It all boils down to the stored energy in the limbs and how much of that energy is actually transferred to the shaft. typically a well designed r/d long bow that stores good energy will transfer more of that energy to the shaft than a recurve bow does. if it's timed well, and has enough preload to stop the string dead you got a winner. Keep in mind.... all r/d long bows are not created equal.
The recurve bow stores more energy in a shorter section of working limb. especially a static recurve or semi static tip. You would think the recurve would be more efficient and have a better cast, and in some cases they do.... But typically a well built r/d long bow in the 50 pound range will out perform a recurve bow of same weight.
The RC limb because of the extra width it needs to stabilize the limb laterally, you end up with a much thinner limb, and vertical stability and limb bulge comes into play on some designs.
The issues that you run into on the RC limbs with lost performance is the parachute effect for one. those wider limbs have a lot more wind resistance as they come forward than a narrow limb does.
The second issue is a recurve limb typically has more mass weight in the tips moving forward than a deep core long bow limb....If you try and use a faster taper rate and thin the limbs down too much, you end up with limb bulge issues that siphon the stored energy instead of transferring it to the arrow shaft. The limb tips are stopping but the working limb oscillates.
This is what is so exciting about using these newer carbon composites and foam. It's making unbelievable advancements possible. We are not only shaving a lot of mass weight in the limbs, but are now able to narrow our limbs up considerably with these carbon composites. We are also eliminating stability issues that are commonly an issue using glass and wood.
These new Static tip RC bows with carbon composites are closing the gap on the hot rod long bows.
Sorry for the getting long winded here guys... but for a bowyer, this is good stuff we are talking here.