With both QuickLOAD and my Excel estimator that works off Hodgdon and Lyman data, I get the estimate that if your gun and barrel met SAAMI test barrel dimensions, you would get about 2834 fps (QuickLOAD) and 2837 fps (my estimator). The pressures look like 45,067 psi (QuickLOAD) and 48,995 psi (my estimator). If your velocity number is accurate (and do try putting the chronograph at least 15 feet away, just in case powder particles from incompletely burned powder are influencing matters).
But I doubt that's the problem. I expect your velocity is accurate because I expect your chamber is a 5.56 chamber or a Wylde chamber that is on the long side of headspace. You may also have a different cross-sectional area than the SAAMI barrel. In QuickLOAD, for example, if I adjust the powder burn rate down 3%, which is the manufacturers lower lot tolerance, and the as-fired case capacity up to 32 grains, which I've seen in from generous chambers, I get 2638 fps, which is lower than your results (so your's isn't as bad as I've seen). Pressure is down considerably in that case (36600 psi).
So what you probably have is low pressure. I still don't know the make of case you are using, but except for a few, like PMP, their capacities don't vary much, with military having the most (contrary to popular rumor, the converse of which only applies to 7.62 cases). Under some slightly unusual circumstances, low pressure can flatten a primer too. One element of that occurrence is having longish headspace. What happens is, the extra headspace allows the primer to back out further than in a tight fit, and if pressure is low, it takes "a while" (by firing event standards) for the case head to start to stretch back to the breech. If the pressure is very low, the case head backs up and the primer cup is strong enough to hold its shape until the head gets back there and reseats it. But you can, in a long headspace chamber, get to a point where there is not enough pressure to stretch the head to reseat the primer quickly, but is enough to start inflating and bulging the backed-out portion of the primer cup if it is backed out far enough. And then, when the case head is pushed back against the bolt face and tries to reseat the primer, it flattens the wider diameter of the bulged portion. About the only way I know to tell the difference between high pressure mushrooming and this kind of low pressure mushrooming is the former usually exhibits cratering around the primer indentation, while the latter has a smoothly radiused primer indentation and no cratering.
So, here's what you need to do to tell which situation you've got. You need to neck size and single-load and fire one of your already fired cases and see if the primer still flattens out a lot or not. If not, low pressure flattening was what you had. To neck size, if you don't have a neck sizing die, just run the case far enough into the chamber to size the neck part of the way, but not touch the sides of the case.
Also, get some CCI 450 primers to try and work the load back up. These primers are magnum primers, but they were formulated by CCI for spherical propellants, which are harder to light that stick powders, generally speaking. It won't hurt anything, plus the 450's cups are formed from 0.025" brass sheet, while the 400 is formed from 0.020" brass sheet, so they have slightly thicker cups and that's another way to fight low pressure flattening.
Again relying on QuickLOAD, even under the best circumstances, 223 CFE is much closer in burn rate to BL-C(2) than to H335. The significance of that is that BL-C(2) is a canister grade (controlled burn rate grade sold for handloading) of WC846, a ball powder developed for 7.62 NATO, and H335 is canister grade WC844, which is a faster burning version of WC846, developed for 5.56 ball ammunition when it was determined that WC846 was too slow for 556 ball. CFE-223's burn rate is between the two powders, but closer to the too-slow WC846. So if I were you I would at least pick up some 62 grain bullets and see if you can't get closer to normal performance with those.
What was said earlier also applies. If you look at Hodgon's data for this cartridge, you will note that the test barrel is 24" long. That's the standard test barrel length for 223. Anytime you see a velocity published by an ammo maker with no barrel specified, you can assume a 24" barrel for this cartridge. A 16" barrel loses about 9-10%, so a published 3300 fps will be about 3000 fps from a 16" barrel. The bullet is spending a bit less time under pressure and being accelerated. Also note that with a short barrel, the slower the powder, the more muzzle flash and blast you get. You might actually turn out to be better off working up to maximum loads of Reloader 10X to cut some of that extra muzzle behavior down.