TL:DR: It’s possible.
It looks like they are running the LED’s at max CREE recommended specs. Assuming that, the XHP35 is putting out 1830 lumens, the XHP70.2’s the other 38k. so 38,170/12 = 3,180 lumens each. If their 40k lumen figure is OTF and not emitter lumens, it would appear they are running the led’s at max factory specs, which CREE claims is 3,800 lumens for the XHP70.2’s and 1833 for the XHP35.
That’s 13 watts for the XHP35 and 29W for each of the XHP70.2’s. Total wattage in that case would be 361 watts. It’s commonly said that between 60-75% of the energy going into an LED turns into heat. Say 60% of the energy turns into heat going into the heatsink, that is 216W of heat. Around 200W+ of heat needs to be dissipated.
So, the LED’s can run very hot, you could run them at 150C Tj continuous, but of course the batteries and the flashlight cannot be allowed to get much above 60C. Decoupling the head from the body with stainless dowels or what not would allow the head to become much hotter than the rest of the light, or decoupling the heatsink more or less from the light is an option. Either way, the target temp for the heatsink can be considered independently.
The XHP70.2 has a thermal resistance of 0.9°C/W, the XHP35 is at 1.8°C/W but runs less than half the wattage so it comes out the same thermally. So if the XHP70.2 is being fed 29W, it is 26°C above the heatsink temperature. Combined with the ambient temp, this allows about a 100°C rise in heatsink temperature before the LED Tj is at 150°C.
What does a commercial actively cooled fansink rated for 200W heat dissipation look like? The MECHATRONICS LED ICE ULTRA is an example of this.
The datasheet says that this heatsink will have a 50°C temperature rise over ambient at 200W dissipation. With this heatsink, at 23°C ambient, the Tj of the LED’s would be 99°C. That heatsink appears to use a 90mm to 96mm fan and consumes 2.76W at 3,000 rpm. They don’t spec it, but it probably has an output of around 55-60 CFM and >0.15” H2O or so. The heatsink resistance is .25°C/W, so this means it could dissipate 400W of heat before a 100°C rise occurs.
What do their smaller heatsink models looks like? Two of their smaller ones are in this datasheet. The difference between the two heatsinks in this datasheet is 10mm in fin length. Their thermal resistances are .58° and .46°C/W between the two. Its about an inch shorter than the 200W heatsink, and the ~90mm fans are 1500 rpm rather than 3000, but the heatsinks are the same 100mm diameter. These fans should be putting out around 30 CFM and .04” H20 SP. The bigger of the two fansinks could handle 200W before a 100°C rise. They do not have tightly grouped fins, presumably so that dust is less of an issue, and this is appropriate for the fans which do not have high static pressure.
If the fans in the X70 are high rpm 30x30x10mm fans, which they definitely don’t look like 40mm fans, they are putting out around 8 CFM each and 0.380” H2O (94.6 Pa) static pressure. Static pressure can be additive as well as cfm, so you’ve got 24 CFM and ~1” H2O static pressure. It’s a decent amount of static pressure; I have a loud and powerful 40mm fan with those specs. This means that the fins of the heatsink can be thin and tightly spaced.
So, what is clear is that the heatsink and fan combination needs to have better than .5°C/W sink to ambient thermal resistance (aka theta sa) to keep the heatsink from rising 100° above ambient, if there is 200W of heat to be dissipated. .4°C/W theta sa would be better, with an 80°C rise.
After looking at a lot of fansinks and specs, it would appear that it is possible that there is a .5°C/W theta sa or better conventional fansink within the volume in the cavity of the head of the X70, especially if it’s copper. (example at .55 theta sa at 26 CFM)(example at .3 theta sa at 26 cfm) The questions are: is there a high density aluminum or copper heatsink of sufficient volume hidden within the cavity of the head of the light? Do the fans deliver 20+ cfm? Are they really feeding those LED’s 361 watts? And so on.