Hey there
I think the overheads are way too shy, you could turn up the volume a bit and also increase (or cut less?) the high frequencies. Boosting your mix around 9kHz with a pretty wide Q, helped immediately both the drums as well as the vocal to gain lots of clarity, so you could improve that easily. The vocals also need some taming around 490Hz to reduce the boxiness.
On the other hand, the bass guitar sounds heavily distorted, but to me in an unpleasent way (check around 1:22) and also needs more automation during the calm part, since the recording has lots of volume divergencies there (e.g. 1:31 vs 1:34).
I do think the vocal delay during the calm part is a nice idea! Maybe you can automate or side chain it in a way, that it's less audible when he's still singing and keep it prominent between his phrases? That may be just me, but during 1:11 I thought someone next door was actually talking to me
An alternative solution might be to pan the delay to the sides, so it doesn't interfer with the main vocal (same with the guitar riff during the intro).
A actually smiled a lot when I first heard your interpretation of the synth part, but in a good way! I heard that song quite a few times now and that really caught me by surprise in a positive, creative way, very cool!!
Hope that helps! Some points may sound like nitpicking, but I think your mix deserves it, cause after adding clarity it really had the potential to shine.
Cheers,
hoshi