I thought over my bumper decision a lot. I was going to build my own out of aluminum to save weight because the available bumpers for my rig ate up a good % of payload, but I too was very concerned with safety. I have a background in risk management so I took a clinical look at it to figure it out. Humans tend to be really worried about stuff that rarely happens, and that causes us to miss out on accounting for stuff that happens a lot and can be way more impactful, so this approach helped me evaluate that a bit.
Risks can be divided into two dimensions: How
probable something is, and how
impactful something is. If something is both super probable and highly impactful, you probably should spend time mitigating that. If it's not very likely to happen, and even if it does, it doesn't have a huge impact, you don't need to worry so much. There's also unlikely, high impact events -- these ones tend to catch us out -- and of course very likely, but not very high impact events. This is a simple risk matrix and using this approach, my process and research resulted in the following:
You are technically more likely to hit a deer than you are to hit another car, based on the data I found. Car-on-car is something like 1 in 300; car-on-deer is something like 1 in 120. So, that's our probabilities, but Impact is the other consideration. The potential impact in a typical Overland context is much bigger in a car-on-animal strike versus a car on car strike. Animal strikes for overlanding usually happen in remote areas, and often on medium- to high-speed roads (i.e. 80 KPH or more). However the forces typically involved mean that you probably won't need airbags to deploy -- a 6000 lbs 4x4 striking a 300 lbs deer will create a lot of carnage, but it won't bring the vehicle to a sudden stop (which is what airbags are good for - when occupants keep moving after the vehicle has suddenly stopped). Without front end protection, even if you are totally safe in the actual collision and walk away unscathed, your vehicle might be dead leaving you stranded in the middle of nowhere. If there are injuries or the weather changes or a travelling companion gets a bee sting and is allergic or any number of other factors bubble up, this can become a life-or-death scenario very quickly. So it's likely, but also highly impactful.
A car-on-car incident usually happens at city speeds (below 50 kph), so less energy in the typical crash, but sometimes the forces involved mean you do want airbags to deploy -- a 6000 lbs 4x4 will be stopped pretty suddenly by another 4-6000 lbs vehicle. Thats why airbag compatibility is important. But, most car-on-car accidents happens in populated areas (which makes sense) so not only are there slower speeds involved, but you typically are closer to things like hospitals, tow trucks, etc. which makes the impact potentially actually much less than an animal strike for my use case.
(Note these are ALL probabilities -- it is of course totally possible to have a car-on-car in a remote area with no help around,
just as its possible to hit a deer or moose within Edmonton city limits -- but it's rare.)
But there are other risks that are way more likely than a deer strike, and yet have a similar impact. Getting stuck is a big risk, and it can have the same impact as an animal strike -- being properly bogged might mean your vehicle might as well be dead, and you are stranded. Again, weather can change, someone might injur themselves in the recovery effort, etc., all of which can dramatically increase the impact of this event.
A well-designed bumper is the most effective way to mitigate these risks. An animal strike with a good bumper makes it more likely that your vehicle can drive away. Getting bogged is always a problem, but a good bumper allows you to mount a winch to effect a recovery, and will often have other options like jacking support or recovery points. A good bumper can prevent your vehicle from excessive damage in a car on car incident as well if the other guy's car and crumple zones are 'forced' to absorb more of the crash energy. His car might be a bit messed up, but yours might be fine. Without a bumper, all other things being equal, both cars will need the body shop.
The odds of getting into a car-on-car accident, or even a car-on-BIG-animal accident, where the bumper would mess with the existing safety equipment of the vehicle enough to change the outcome for your passengers is pretty small, but its still a risk. You can make it smaller still by going with an ARB, because they do test for exactly this.
So, I have an ARB on order*. Thank you for attending my TED talk.
*(Technically I'm gambling a bit -- the bumper I ordered is for my truck's twin, the Colorado, but I'm going to see if I can make it fit my Canyon (which it should). This might reduce that airbag compatibility but I don't think it will because of how similar the Twins are, but either way I'll be the guinea pig for Canyon owners who want the ARB Bumper!)