Ok. I listened to the acapalla closely this time. A few things:
-Pay attention to consonants (S, T, K etc.) and use volume automation to 'smoothen' them. I heard a few T's that are too hard and I heard the S as well somewhere going to hard. You can even cut out these 'hard' sounding consonants from the BGV. If the consonant is in the lead(s) then it is often enough. Sometimes production choose to turn a T into a D, to smoothen it (ariana grande - into you for example), so that's another thing to learn
*now you could use a de-esser for the job, but I recommend trying to solve it with volume automation at first and not merely rely on a de-esser because it quickly tends to do more harm than good for those not experienced.
-Timing: Currently the timing isn't tight enough as getawaydriver already mentioned. Make sure that at least the BGV have the same timing. You could use software like revoice pro (also known by the name of its younger brother: vocalign) or you could pay very close attention to your vocalist being right on the timing during the recording session. Since this one is already recorded, use revoice pro to fix the timing. Use the lead vocal as the reference input for the program and align the rest of the vocals to the timing of the lead vocal. The lead vocal itself can also be aligned manually, that means chopping audio, using volume crossfades to get the timing as pleased.
-Panning: I gave you a tip on panning last time. It is important that when you pan for example the lowest voice to the left, that you have a double of that lower voice on the right. Same audio file or another one? Try and see what works.
-Delay: your delay sometimes pops up in the middle of a vocal line. After all, this will disturb a mix. Control your delay manually or use a ducker or a sidechain which ensures that the delay is only heard when silence reaches. At any cost, those delays heard sometimes in the middle of a line need to go out. You could also print the delay manually and then later cut it out. There are different methods. Another method is automating the send input in case your delay is on a send. It should be on a send though in most cases!
-Some plugin tips: If you are not an experienced mixer, make use of an assistant! The bad news is where to find one is a shit task but the good news is that AI, as found in Izotope Neutron (mixing) and Izotope Nectar (mixing vocals specifically) have an AI mixer which sets up certain stuff for you which you can tweak afterwards if you don't agree with its suggestion.
-Can you tell me a little about the mic chain? In what room is the recording, what mic, where is the cable of the mic going into, what hardware is doing work and what plugins are on the chain?
-Harmonies: If I'm not mistaken all the melodies of your harmonies are exact repetition of the lead but on a different pitch. This is a mistake. When you study danthemanbardo videos, you will see - particularly in the advanced harmonies which the BSB made - that they aren't always exact. 12tone has a video about harmonies as well where he discusses several types of harmonies. The melodies of the harmonies often are different than the lead melodies itself and that's where the real magic happens, that's where u get goosebumps hearing the melodies. Creating harmonies of groups requires very intelligent execution and my best advice is just to check out those danthemanbardo videos for hours and hours and find patterns, that will give u the best insight in what decisions are logic and what decisions are not. After all, as Aristotle concluded, 'excellence is never an accident. It is always the result of high intention, sincere effort, and intelligent execution; it represents the wise choice of many alternatives - choice, not chance, determines your destiny'. Also in the art of creating harmonies excellence requires to make the right decisions among many alternatives.
Good luck
