Speaking of CNO fusion and fusion in general. The Sun is indeed a low power density fusion reaction. But, again remember that it is burning by the P-P reaction (except ~ 1 % by CNO). If it was burning D-D it would be ~ be fusing at a rate of ~ 10^15 times faster. If it was burning D-T it would be ~ 10^17 times faster.
In a heavier star the CNO cycle scales with temperature at ~ the 17th to 18th power. At seveal hundred thousand eV this could max out at ~ 1/1000 to 1/10,000 as fast as D-D fusion at similar temperatures. I am not sure how this compares from a density standpoint. With the density considered the fusion power output may actually be 100-1,000 times higher than a Tokamak burning D-T. In reality massive stars may be burning hydrogen by the CNO cycle at a rate ~ 10,000 to 100,000 times faster than the Sun. This doesn't consider the volume of the star or it's core so the output per unit volume of the core or total stellar volume to the photosphere is unknown.
If 15N or 15O (I think) is used without utilizing the slower CNO steps, the fusion rate could be ~ 100 times greater, or if my memory serves me ~ 1-2 orders of magnitude smaller fusion cross sections compared to D-D cross section at the same temperature of several hundred KeV.
In a reactor burning p-15N at several hundred KeV the fusion rate would be ~ 1/10 to 1/100 times that of D-D. This occurring within a massive star that somehow had an abundant supply of 15 N would burn with impressive fusion energy output. A massive star may burn through the available hydrogen (perhaps as much as 10-20 times the amount of hydrogen in the Sun's core) within ~ 1-10 million years by utilizing the CNO cycle. By utilizing the p-15N reaction without the rate limiting (and carbon recycling) slower steps, the star would burn through it's aviable hydrogen in as little as ~ 10,000 years. I don't know what the volume of the giant's core or total stellar volume would be, but I suspect a few cubic meters of this imaginary core would compete with terrestrial hypothetical fusion reactors- possibly beating Tokamaks, matching Polywells, and approaching DPF, and that is with hydrogen (and 15N) fuel. If D-D or D-T fusion was occurring in this starthe fusion energy density would be at least several orders of magnitude greater yet.
This addresses my pet peeve that many consider stars as feeble fusion reactors compared to terrestrial reactors. The comparison is meaningless unless fuel considerations are included, along with temperature, density and system considerations. For instance a Tokamak plasma might have a given volume, but the system including the magnets, shell and building all would need to be considered if you are comparing it to a star with it's reacting core and all of the overlying non reacting layers.
The biggest problem with 15N is Bremsstrulung X-ray radiation. As Bremsstrulung scales as the 0.75 power of the temperature and the square of the Z . Nitrogen with a Z of 7 would result in increased Bremmstrulung of 49/25 or ~ 2 times that with Boron. This does not consider the effects of excess hydrogen in the fuel mix, so the net effect may be more subtle. Also, the P-15N fusion cross section may be several orders of magnitude less than the P-11B cross section at similar temperatures. With an optimistic Q of ~ 5-20 for P-11B in a Polywell, I doubt P-15N could have a positive Q. But, if Dense Plasma Focus x-ray recovery methods are utilized (perhaps 80-90% recovery efficiency) perhaps some net useful energy might be squeezed out in a larger machine. But as other losses would increase with machine size, it might also require some other fusion magnifying effect like POPS.
It would be sexy if isotropically purified ammonia or even water could be used as aneutronic fusion fuel. Of course the same could also be said for D-D fusion (from water) if you don't mind those pesky neutrons, and it would probably be several orders of magnitude easier- relatively easier but still unabtanium thus far.
Reflections on fusion history, current events, and predictions for the 'fusion powered future.