I'm late to join in here, but I'd take a different approach.
In my experience, your microphone sensitivity, low cut, and recording level (manual or auto) are all very interactive with these Olympus recorders. Some examples of what I'm trying to describe follow:
First, let's cover low cut. There will never be anything desirable below 100 Hz when we're talking about human voices, and in terms of environment sounds all those low frequency rumbles do is mask the good stuff. In addition, if you're using auto for your recording levels, all that low end junk will be triggering the in built limiter on your recorder, so even when you cut those frequencies in post production you'll have some work to do to counteract the audio ducking. My view is that if you know for sure you're gonna get rid of those frequencies afterward anyway, don't bother bringing them along for the ride at all. I mention 100 Hz specifically as that's usually the least dramatic low cut option on Olympus devices. Tbh, unless you're recording through something solid (walls, doors with no discernible gap etc), I'd probably ramp it up to the 300 Hz setting, especially if you're working with the combination of the high microphone sensitivity and the auto recording level setting. When you've got that combination engaged, the less masking at the source the better.
Mic sensitivity - If it's a file that I don't intend to do much work on afterward, then sure, high will do a good job. If I know I'm going to put work in to clean things up and improve the audio, I usually go for middle. You'll get better results amplifying with a computer to crunch the underlying numbers and some meters in front of you if you like to work visually. I find that middle mic sensitivity gives me a good compromise between leaving myself with some head room but not having to do a ton of work fighting against the extra hiss that comes from amplification.
Recording level - A lot of what I wrote above hinges on this setting. As much as I hate the sound of auto-gain and similarly named functions, I still tend to roll with the punches and keep this one on unless I'm sure that the environment isn't going to throw up any surprises. In a straight choice between the sound of auto-gain and the sound of clipping, I'll take auto-gain every time. Distortion is bloody hard to get rid of, and even the pro plugins only do a half decent job to my ears. I find that backing the mic sensitivity down to middle means that the onboard compressor/limiter isn't going nuts, and the sound of auto-gain at work doesn't drive me quite so crazy.
Mic zoom - so far, I've heard nothing good come from this. As I understand it they use tricks with phase to simulate movement or changes in the arrangement of multiple mics. IMO, you don't want to fuck about with phase unless you know you're going to gain something from it or you've got a specific repair to make. In my tests, it didn't sound particularly convincing. I'd sooner switch to mono and/or use my body to block out sounds bleeding in as much as possible before I went with mic zoom. Really, if you want a super narrow pickup pattern, you're better off buying an external mic that has that.
Recording mode - unless storage space makes it impractical, always record in PCM, Wav or whatever your recorder calls it's uncompressed format. If you do any work on the file between recording and uploading here and you've recorded as mp3, you're going to re-encode the edited file before uploading. Stacked encoders sound bloody horrible. For the kind of stuff we're recording here, so long as you're careful with levels I don't see any advantages to going higher than 44.1 KHZ PCM, especially when you consider the amount of storage space the extra resolution would burn through.
So to sum up, my starting point for a recording like this (and most others) with an Olympus LS series device would be mic sensitivity on middle, some form of low cut (100 for sure, 300 if you're feeling brave and/or arent' heavily into your post production yet), auto for recording levels unless you know the target and surroundings well, and recording mode set to wav. If you've got no post production chops or you just want to get the action straight from your recorder to our eager ears, then mic sensitivity on high and the dreaded auto-gain will spit out something much closer to finished sounding.
I realise most of this completely contradicts what Emmpet8 has been saying. On paper he's not wrong, and when you're working with a consistent source, decent mics into a quality preamp then his advice is right on the money. Even if you're not working with any of that (these field recorders aren't really any of those things yet), his approach is still great so long as you've got the patience and skills to tidy up afterward the way he can. I have neither, so this is how I get usable results up up here with the least amount of time spent. Of course, as the quality of the components and DSP in the field recorders improve, this stuff is going to become less relevant. It's also worth keeping in mind that we're talking about small differences with most of this stuff. Yes, they do all add up to something significant in the end, but you've already crossed the threshold in terms of hardware quality where it's difficult to make a "bad" or "unfixable" recording, so don't drive yourself insane over these details. The more you use it, the more you'll get to know your device.
BTW, the actual audio in this thread is a great catch!
Sc0tt-UK