I don’t know what’s your own opinion about the progress you made so far, but to me, as a listener, I think you improved a lot. I remember older tracks posted here for feedback that where sounding quite muddy, with a lack of clarity & definition and suffering from frequencies masking. This has changed a lot since in my opinion. Maybe it’s time to ask yourself about your monitoring now. I know that most of us don’t have the luxury to build the perfect setup in the perfect room, but it’s important to find a solution that at least you can “trust”. If you can’t do otherwise than mixing on headphones I think that getting a “trusty” model really helps. Models like the Beyerdynamic DT-770 Pro 250 Ohms are very affordable now, the only draw back with such closed-back headphones is the lack of stereo imaging and lesser space definition, open headphones like the Sennheiser HD 600 are better for this, they belong to older models from the brand now, so they are also less expensive than before and still very praised. One or the other coupled with Sonarworks Reference is defintiely something that will help a ton. Adding monitors is also something to think about, I won’t see myself mixing 100% on headphones ( I used a closed back model ), I like to have that notion of space & stereo imaging and I find listening on monitors is the right solution for that. Have you heard about the brand Kali Audio and their LP monoitors series ? They are very affordable and seem to be a reliable solution, and again used with Sonarworks Reference, it can drasticly changed the way you’re monitoring and what you’re listening.
I’ll be honnest, it’s kind of weird to start using Sonarworks Reference, but it’s a matter of time and listening to tracks that you know well and you’ll then start to see the benefits of it.
For mixing thecnique, well there’s no secret, it’s a kind of art and it takes a lot of time and practice to get it right. What I think is good to keep in mind is the audio workflow in any DAW : first is your source, then your effect chain on the channel, then your channel fader and finally your master fader. Many times it’s not enough to adjust channels faders to get a good balance, it’s sometime more efficient to act on the source gain ( sample level, soft-synth output ) and it’s also good to use a gain utility on the channel instead of tweaking too much your channel fader. DAWs faders are based on logaritmic scales and they also tempt to measure peaks levels rather than average levels such as RMS or loudness.
Check this video.
So in the end they are more accurate when used around 0 dB, if you start to have your DAW faders all over the place accross channels and with huge values either up or down, then there’s somethng wrong with the source material level, so that’s where adjusting the source level and also using a gain utility before the channel fader makes sense.
And there’s also other things than just levels, for synths for example, as your friend mentioned “too much synths” it can simply be that the sound is already too huge, taking too much space in the mix, and many times it’s just a matter of closing the fliter or reduce the spread or the intensity of the synth plugin built in effects like chorus, delays, compressor, reverb…
I think I already mentioned this, but do check Phil Johnston’s course on Fundametals of Mixing
For the arrangment part, progression and things like bass line and drums parts variations, it’s mainly something you’ll get better at if you use reference tracks and if you take the time to deconstruct many reference mixes.