Note also that you pay upfront for acquisition but your revenue base builds up over time - the more momentum you get, the better off you are.
It is however true that a lot of businesses that rely on paid acquisition end up with low margins - just look into the leadgen industry as a good example of this. This is particularly true in the Google Adwords world where auctions systematically drive up prices.
That said, if you can pay $80k and get $120k out, and that scales up to LOTS of revenue, then making $1.2B in revenue per year on $800M in cost may not be that bad. Certainly better than a lot of businesses.
a 50% ROI seems pretty good compared to traditional businesses
If a business spends $800 and makes $1200 that is great. The risk is small. If the same business spends 8k to make 12k that is great. The risk is bigger but relatively small. If the new hired gun, self-proclaimed-marketing-guru, decides to jack up ad spending to 80k to make 120k, someone is betting too high now.
Asking if this model is scalable is bringing emotion into this scenario which will guarantee you failure sooner or later.
At one point you will reach the amount at which spending more on ads will not increase your revenue. I guess that is when justifying "branding" comes.
One of the biggest examples of this is eHarmony, which is rumored to have $100M+ revenue, but terrible margins because they have to plow so much of that money back into advertising.