CFM ratings are not actually done by measuring the volume of air moved in 60 seconds, they take the velocity of the air exiting, and multiply it times the diameter of the aperture. There is NO restriction to airflow in this test, and when there is actually restriction, the actual amount of air moved can be a tiny fraction of that CFM claim.
Even if there is no restriction to flow in a certain application, most fans have hotspots of flow exiting the fan. No doubt the anemometer is placed to measure these higher velocity areas of flow, making the published cfm figure even less representative of real life efficacy.
Computer fans use a static pressure rating in addition to cfm rating, and impeller blade design has a huge effect on this outcome. When attached to a computer heatsink a low cfm 120mm fan, like the Noctua NF-F12@ ~ 53cfm, but which has a high static pressure rating for its 1500 rpm, greatly outperforms fans with 70% higher CFM ratings and which also draw 2 and 3x the amperage, and create more noise while being less effective.
The 100 cfm duct booster fan, with its flat metal blades, relies solely on angle of attack of those blades to scavenge and push air. Such blades can be easily stalled when enough resistance is placed to the airflow, and even a smooth duct presents a surprising amount of restriction. These fans work best when the air they are moving already has some momentum from another fan.
The computer fans blades are foiled nicely and rely not only on the angle of attack but this airfoil to efficiently move the air.
The centrifugal fans like the Broan unit are basic and effective, but are wasteful of electricity. I have one similar looking design, and while I cannot easily measure airflow out of the 3 inch vent, I did measure amp draw, and at 77 watts, decided against employing it to scavenge air from my workshop ceiling. Seems this centrifugal design is much less affected by restriction, and is perhaps why your 50cfm broan fan is outperforming the 100cfm duct booster in yuour subjective impressions of efficacy.
The amount of restriction out your rafter is likely an area for improvement of efficacy.
As I want heat evacuation without relying on the power grid, and with everything locked up and me gone, and I have solar and some batteries already set up, I employed some 120 and 140mm 12v computer fans, and a cardboard divided housing which forces their flow out a rafter at roof apex.
The 120mm fan is rated at 74cfm and 0.24 amps and does actually draw that much. Some computer fan ratings are the start up surge, like the 140mm fan next to it, which claims 7 amps, but I've only measured just over 5.5 on the start up surge. I have put this 140mm fan on its own switch, and control the speed/ voltage via a modified voltage bucker. Flat out at 12.8v input, this 140mm fan draws about 3.25 amps and turns 4k+ rpm and is very loud. If placed face down on a flat surface will levitate about 1/4 inch over it. Its 7 large wide blades jut well forward and are a low angle of attack, for a very high static pressure rating, but considering the noise made and amp draw, it has a comparatively low cfm rating. I have another 140mm fan with more fan blades, shorter and higher angle of attack and a 2.3 amp rating, which in free air moves significantly more air at higher velocity, but it cannot levitate itself, the blades just stall.
I only turn on the 140mm fan at generally no more than 0.7 amps of draw via my voltage bucker, come afternoon, but the 120mm fan is fed battery voltage directly, 24/7, and when I open the workshop door late morning, it is still cooler inside than out, whereas before i added this 12v exhaust shroud the opposite was true.
The combination of radiant barrier and exhausting the ceiling can be taken to unneeded extremes, but its nice to not bake in one's own workshop on the hottest of days.
I want to remove the grid entirely from my workshop ventilation, but the 20 inch box fan with 20x20 suction from ceiling shroud is hard to beat in terms of effectively replacing interior air with outside air. Placing it on an inverter at high speed is ~430 watts though, and I only have a 100 watt solar panel feeding my 190Ah of Deka intimidator AGM.
In terms of efficiently exhausting air, for low wattage consumption/ noise, the computer fans are hard to beat, but 230mm seems to be their largest size, and those in this size range are just low rpm designs and many would be needed.
There are 115VAC computer fans, but much less options than their 12vDC counterparts. I think even if I did not have batteries to power my ventilation I would be using the 12v fans on a transformer/ switching power supply as the amount of design and thought that goes into the computer fans by companies like Noctua, Delta and others, makes fans like the inline duct booster fan into a joke of design.