I was thinking earlier about landing with suicide burns. I know they're not as efficient as constant altitude landing, especially at lower TWRs. But then I got to thinking.. the Oberth effect makes a rocket burn more efficient effective(?). Constant Altitude burns generally feel like they take longer, perhaps because you maintain a low altitude while you kill horizontal velocity, which suggests to me that as they decelerate over a longer period of time, the net Oberth effect would be lower. Otoh, a suicide burn would accelerate you to the highest speed before you hit the LFO brakes (and hopefully not the lithobrake, unless your experiment is to find out how big of a hole you can make - which would be wicked fun in a destructible environment).
On a scale of something like the Mun or Minmus it's not really relevant, but for something like Moho, where you come in nail-bitingly fast with no atmosphere (or at least... I did the one time I tried - my landing attempt turned into a very fast low altitude flyby followed by many years of floating around in space lol) if you had a sufficiently high TWR to do a meaningful suicide burn from a transfer whose periapsis barely impacts the surface, would it exceed the efficiency of a circularization and constant altitude landing?
It would probably never be more efficient if you look at the whole mission (for example, you'd need to heave around a bunch of very heavy engines to try and dump out 3000m/s or so in 20-30s which would be more costly to send in the first place) but it's still a question that came up. Also using something like mechjeb with a sucide burn timer so you don't bugger it up (which you'd have little room for error at those speeds).