the aroma they give off will evolve and change several times between seed and smoking some cured weed. The less agitated or stressed they are, the less smell - typically. genetics will play a large factor. Possiblities are too varied to give a specific generalization about this. using info from different gardens and perspectives... gonna be some fog of war for sure.
300-500 is a bit low, assuming it is also in proper ratio/balance. If ayou dont have much time left... meh.. flip a coin it won't matter what you do with so little time left. in soilless/hydro a 750ppm/1.5EC medium is a good ballpark to start in. that will keep a plant churning at full bore. Soil is a bit of a translation from that... what came with the soil will change how that is implemented. you can feed higher, but then mix in water-only irrigations between too... but a constant mix is going to have better results. it'll be in a more optimal range for longer periods of times and therefore more optimally used... studies with frequent irrigation support this extrapolation. it's why 'hydro' typically outperforms soil even without multiple irrigations per day.
-----
flushing is a myth. it doesn't impact mineral content of buds in any observable way. All the 'old' reasons the culture gave for flushing are objectively wrong - this is not an opinion. it is fact. unless they made up some new reasons, lately? but that in itself is a red flag they know not what they speak of... can't change logic midstream and maintain your integrity, lol.
You flush when your substrate is toxic... to lower overall EC/PPM. It is a tool used for a specific problem... not harvesting. This is a learning lesson -- popular beliefs with no studies to back them up are often bullshit. Anecdotal evidence is sketchy for this exact reason (lucky if it is weak to mild correlation with reality, or in this case, a correlation coefficient of "0", lol. These ppl that perpetuated this myth had virtually no understanding of biology and lacked the tools necessary to even know what was going on in a resolved way. So they made blind extrapolations with little relevant knowledge... what could go wrong? A lab can measure this stuff in a precise and accurate manner (the equipment has been around a long time). Blind taste tests show it's bullshit too.
Flush the last week if too lazy to mix nutes or cause soil ppm is high enough to maintain a healthy plant. The latter i'd just water normally with limited runoff to ensure no dry pockets, cause flushing invlolves excess runoff... which is only needed to correct improper concentration levels. The loss of yield will be imperceptible to a human eye, but it is quantitatively a net-negative effect when splitting hairs.