If light in the range of ~420-430nm were required for coral growth we'd be seeing a lot more 420nm actinic tubes augmenting halides in the days before LEDs. I just don't see any evidence that light between 400-450nm is mandatory for coral growth other than manufacturers trying to push exotic color combos to differentiate themselves from others. From a biological perspective blue light of any wavelength is sufficient. At least that's what the pHd's say.
First off, there are an enormous array of actinic T5 lights that are and have been available
FOR YEARS that have been specifically geared for irradience below 450nm...like the geisemann pure actinic:
And anyone who has pursued good coloration with sps and acans knows that this is a damn near mandatory bulb, in association with other actinic varieties from other brands with a dominant spectrum between 400nm and 430nm.
Secondly, The necessity of sub 450nm wavelength light has never really resided in the realm of growth requirement, however there is equally no study to disprove that the use of sub 450nm light is
NOT beneficial...so your assertion holds equally no weight.
And no, the specific blue light is specific to each coral and the structure of its PCP makeup...that is perenidin and chlorophyll, where chlorohpyll a peak absorption is at 410nm and 430nm respectively, but PCP which is found most in corals peaks in a variety of wavelengths...
so to say that any blue light is sufficient while anything purple is not required because you don't think it is...is an understatement at best.
Truth is, the only frequencies missing from the older generation leds compared to MH's has been the sub 455nm light...which led to the inquiry of whether those frequencies were needed to have a successful LED setup for sps corals.
Thus far, it has been a resounding yes, from nearly every single successful LED user. The unsuccessful ones just cry and go back to their MH...even though the lighting wasnt the problem.
The dominant blue wavelength of single color and white LEDs is far broader than halide or fluorescent tubes, not the other way around. Sanjay Joshi's spectral graphs show just how narrow the spectral range of halides actually is.
I don't know where you are getting this, but MH and T5 cover basically every frequency in the visible range including a lot in the non visible...they cover an number more wavelengths than leds.
sanjay's graph
LED spectral plot:
you can see that the leds dont cover as much in the sub 455nm and the post 660nm as the MH do.
LED fixtures have unique potential problem I've mentioned in other threads, and that's optics. Optics collimate point light sources and are likely the cause of the problem. Do you use large fresnel lenses on your halide lights to focus light and increase PAR values? No? Well then...don't use optics on LEDs.
This is a valid point and should be considered by novice users as a potential source of any problem with their new leds. However, for advanced users it tends to be a near non existent issue...because the lights are of the correct variety, dispersion, and intensity to avoid this as an issue.
My assumption is that the LEDs narrow wavelength spectrum happened to miss the wavelength that particular coral needed. Therefore, it was as if no lights were above the coral at all.
Having said that does anybody know of any research done with LED's and MH side by side.
There is a plethora of research, most of it is linked to within this thread itself. Do some digging! But to address the "narrow" thing explicitly, we are lucky int he fact that we know what these frequencies are, and all of the leds today can meet these requirements....this is why so many of the advanced user have had enormous success with leds.
If LED's are proven to be as good as other light sources (which I think they will be, the possibilities are ENDLESS.) I think they'll become more popular and prices will come down. Metal Halides are just overkill for most systems, while extremely effective, overkill IMO. Personally, for the price/area effectively lit, I would go with metal halides, for now. 
