This whole astrophotography thing is quite complicated in one sense but in another sense it really isn't. One of the brilliant things about mathematics is that as long as you're consistent you can do whatever the hell you want! Let me explain.
When I am comparing optical systems to get a sense of SNR and integration times I calculate a rudimentary flux term for comparisons which is telescope aperture/resolution. Take the FSQ106 KAF-16803 as an example.
106^2*((9*206.265)/530)^2 which gives me a value of 137,847 (I'm rounding to whole numbers).
Now let's compare that to a system that has near identical image scale but done at 5% of the cost!
QHY183M + Samyang 135mm F/2 @ F/2.8
48.2^2*((2.4*206.265)/135)^2 = 31239
They both have very similar FOV, very similar image scale of 3.5"/pixel (FSQ) and 3.67"/pixel but the FSQ is 4.4x higher SNR for the same integration; 5 hours with the FSQ would require 22 hours with the Samyang for the same SNR. This SNR calculation is purely for extended sources like dusty regions as point light sources are purely a function of aperture where the FSQ captures 4.8x the amount of light.
With the Dragonfly array it depends on how you look at it. One one hand you can say that an increase of 30 to 48 400mm F/2.8 lens' is taking you from a 782mm F/0.51 to a 990mm F/0.4 OR you can think of it as being able to capture 180 hours per night to 288 hours per night (taking 6 hours as being your imaging time on a spot of sky at a time) with a lens. It all depends on how you look at it.
When it comes to professional observatories and large apertures, you're generally chasing down significantly fainter objects where capturing large amount of photons does become important. Take what Suavi mentioned earlier when he said he was capturing as little as 1 photon per hour in some patches around the Helix. Comparing this to Keck and its 10m mirror which would be capturing 2.52 photons per second (9070 per hour) from that same patch of sky, you begin to understand why bigger can be better! Now Keck is imaging at a much higher image scale than Suavi so those photons do get spread out a lot more amongst the pixels so it may end up working out to more like 300 photons per hour per pixel but with the added resolution that goes with the increase in image scale.
EDIT:
I've had it pointed out that my figures don't quite match what they should and I've figured out that I should be multiplying the aperture and not dividing it.