I voted #3. Honestly, the only reason I can vote #3 and not #2 is because a few years ago I spent a significant amount of money acquiring a huge number of natural essences ( > 150 different oils/absolutes, etc ). This let me learn what the oils and extracts smelled like in isolated form, although an important caveat is that often these extracts do not smell like the living plant and hence may not be an accurate representation of the note in question. Rose absolutes, for instance, do not smell like a living rose, instead smelling more winey, peppery, decayed, with hints of pickle, etc (qualities heavily dependent upon their location of cultivation and the quality of the extraction itself.)
That is why, when I do my detailed breakdowns of scents, I will usually say, "this smells like X oil" if I find it smells like the oil moreso than the note in nature. In Reflection Man, for instance, I get a very strong orris butter note presence, which is different than the scent of the iris flowers (the ones that actually have any scent). To break it down even further, Reflection utilizes an orris with a lot of ionones which give it a very sweet and shimmering effect, with undertones of raspberries or even hints of grape.
Despite knowledge of individual notes, anyone with any blending experience realizes how quickly notes get lost in the mix, and how two (or more) notes can blend together seamlessly into what appears to be an entirely unique third (and only) note. A tiny bit of fennel seed oil + ylang ylang extra + lavender oils + absolute + a hint of jasmine and beeswax absolute can combine to make a fairly convincing lilac accord (ok, a fairly basic lilac accord, but a start).
Creating new or novel accords is interesting and definitely a large part of perfumery, but I think the true art lies in the way accords and phases of development are woven together. It is very very difficult to have three or more distinct stages in a perfume with seemingly disparate notes and yet to unify the composition seamlessly. It can even be very difficult to make a seemingly linear scent, depending upon the aromachemicals present for the accord you wish to achieve. Creating a long lasting lemon note, for instance, may be very difficult if there are no lemony basenotes available to a perfumer (i'm not aware of all the aromachemicals available so I don't know if this is actually true, but it seems like it must be as there are few long lasting lemon scents). Instead, a bit of illusion and trickery is required. If you can get a note that is "lemon like" (eg: immortelle oil has a lemon like note along with its tea-like characteristics. Keep in mind I'm talking about the essential oil steam distilled from the immortelle flower and not the absolute which produces the maple syrup like note) and then utilize some other notes to "hide the seam" where the lemon notes (say lemon, litsea cubeba, and lemon verbena) fade out, the wearer may perceive this as one long lasting lemon note that is just changing ever so slightly, while in fact all of the lemon oil is long evaporated.
The amount of balance required to create interesting scents that are not too sweet, too cloying, too dirty, etc. and that have no notes jutting out in truly unpleasant ways (but perhaps have a note here or there jutting out just a little to create some interesting tension or contrast) is truly very very difficult. Perfumery is hard. Perfumers should be respected!
I've learned a LOT since working on my own perfumes, and realized just how much there is to learn. It's both daunting and exciting. This is a hobby that will remain interesting lifelong. If you start to get bored with the hobby, consider delving deeper down the rabbit hole. You have no idea just how deep it is!