I love that game, The Stanley Parable. I played it once, but that session went on and on till I found as many endings as I could.
Here’s the situation as I see it. Your quality is already comparable and surpasses that of most streamers. If that’s all you were going for, then you’ve achieved it. But if you want to really push the limit and set new standards for that industry, you can.
The roll-off on the mic is a great feature to be using as it’ll remove any of the surprise vibrations that make it past your shockmount and the towel you’ve placed under the mic stand. That’s taking care of any sub-bass rumbles, but you’ll still benefit from some post-processing (which I don’t think you’ll want to do since you’ll have to decouple your voice from the stream and then re-render everything, not gonna work for Twitch). So you’ll want to do some processing live in the moment. One thing you’ve not mentioned is Audacity’s equalizer: http://manual.audacityteam.org/man/equalization.html
I bring up the EQ because, like most everyone is forced to do, you’re recording in a small room, likely close to a wall or corner, and likely with either zero acoustic treatment or not enough. You don’t have a fluttery sound or a lot of reverb, but you do suffer from some typical bass region problems due to lack of corner mounted acoustic treatment. You’ll be able to see drastic improvements with the equalizer. And if you don’t ever move yourself from that spot, you can create a single, set-and-forget EQ curve to use each time.
I wouldn’t be afraid to turn off the roll-off on the microphone and do it all in the EQ, and I’d test pushing the roll off as high as 150 Hz up to 200 Hz, so there’s a slight roll off up there but by the time it hits 60-80 Hz and below, you’re chopping out all of that from your voice. It won’t sound thin, but what it will do is create a separation between your voice and the low-end of the gaming sounds, which is going to be vastly important. But for the lack of acoustic treatment I’d do a wide dip and test cutting up to 5 dB out from the range of 250 Hz to 500 Hz. That’s usually where you hear the bad effects of ‘not enough acoustic treatment.’ That will bring a lot of clarity to your mid-range and upper frequencies. Finally, I’d test a small boost of 2-3 dB around 2 kHz up to 5 kHz to get more presence and intelligibility out of your voice. This will let you keep your voice at a comfortable volume level while still being heard clearly over the gaming sounds. People who suffer from this issue compensate by turning up their vocals, but that’s not what you want. You want balance.
You’ve got to realize that you’re not just working with your vocals, you’re basically mixing your vocals into a pre-built set of music and sound effects. If you can do some EQ work on the video game sounds too before combining the streams, then you can start notching out slots there for your voice to pop through, like cutting at the same spot you’re boosting in the 2-5 kHz range. This is a good basic discussion of all of the tactics you’ll want to be using: https://ledgernote.com/columns/mixing-mastering/mixing-eq-tips/ You actually don’t have any sibilance or plosive issues, which will save you a lot of EQ / compression grief. That mentions an advanced compression technique for ducking that you may or may not want to use. I probably wouldn’t but you could drop the volume of the gaming side by even 2 dB whenever you start talking, which could help, but if you do it I’d make it very unnoticeable. Probably no more than 1-2 dB.
I wanted to mention basic compression though, in case you aren’t using it. You don’t have a lot of crazy variances in your volume, which is what compression solves, but it never hurts to make sure you’re applying even 3-5 dB of compression regardless, which will bring up the volume of the quiet portions. Once you’ve solved your noise floor issues with the computer sounds, etc., this will be a huge boon to your listeners. I’d make sure to use a compressor AFTER the equalizer and set a fast attack and fast release with a compression ratio of around 5:1, then drop the threshold till you’re seeing 3-5 dB of gain reduction. You can then add that “lost” volume back with the output gain. But if you do this after you set up that perfect EQ template sucking out the bass issues and adding intelligibility and presence, you’re going to have a crystal clear sound. You may even consider a slight high-shelf starting around 8 kHz to add some sparkle and air to your voice. This will help you cut through the mix too.
Good luck, I hope this was helpful.