Livermore Valley AVA is part of the San Francisco Bay AVA, part of the Central Coast AVA. I really like this geographically hierarchical structure of appellations. It's simple, logical, and helps the consumer properly figure out where the wine they're drinking is from. The details of the AVAs don't matter so much - what matters most is their Russian-nesting-doll approach.
In Provence, an area I know quite well, there's AOP Cotes de Provence. Then there's five DGC (dénominations géographiques complémentaires), which aren't subdivisions or appellations in their own right, they're just a detail (La Londe, Fréjus, Sainte Victoire, Pierrefeu, Notre-Dame-des-Anges). This is separate from Coteaux Varois en Provence AOP, Coteaux d'Aix en Provence AOP and Les Baux en Provence AOP which all have the Provence name but aren't subdivisions of Cotes de Provence so much as they are siblings. Then there's IGP Alpes-de-Haute-Provence which has the Provence name but is an IGP not an AOP and makes mostly reds. There's also a bunch of other appellations in Provence but without the P-word in their name: Cassis, Bandol, Palette, Bellet (which isn't in Provence at all really), Pierrevert. All of these can be declassified to IGP Mediterranée (which theoretically but not in practice includes Corsica), and some alternatively to IGP Var, others IGP Maures, some IGP Pays-des-Bouches-du-Rhone, and some to IGP Alpes-Maritimes, IGP Alpilles, and of course IGP Alpes-de-Haute-Provence.
This is starting to get mighty complicated, and I think we've got to 21 ways of saying "it's from somewhere in Provence", before even talking about whether we're on clay or limestone, slopes or plateaux, or whether the wine shows any particular style or terroir. Does the customer really care whether a parcel was one side or another of an arbitrary administrative division, as long as both sides share the same macroclimate, soil, grapes and style of winemaking, or in this case, 'cultural region'?
Let me paint a picture instead of my dream appellation system in Provence. We'll start by keeping the IGP Med for pretty much any wine of any grape (within reason) of any yield in any style that happened to grow locally. This is really for inexpensive vin-de-table or whacky experiments, and I would expect very little to be made. We'll ditch the departmental IGPs, they'll either declassify to IGP Med or be bumped up to an AOP. Like Bordeaux and Burgundy, we'll have a generic 'Provence AOP' that includes basically the same geographic space as IGP Med, and for which basically any vineyard in PACA that wants to can qualify. We'll keep it to traditional grapes, and a moderate yield (maybe 60hl/ha), without too much quality control and very little typicity control.
The next step is the sub-appellations. All the DQCs get their own sub-appellation. AOP Frejus, or AOP Provence (Fréjus) sort of thing. The non-'core' Provence appellations (Bandol, Cassis, Palette, Bellet, etc) get the same treatment, so in practice AOP Bandol continues, but with ability to declassify to AOP Provence, and the right to use their own weird grapes (Folle, Bracquet, Tibouren, Marsanne, Clairette, Bourbolenc, you get the idea). Coteaux Varois and Coteaux d'Aix get the same, being AOP Provence (Coteaux Varois) or AOP Coteaux Varois, or even AOP Coteaux Varois en Provence as they have now, but as children of CdP, so with the ability to declassify to AOP Provence or IGP Med.
From 21 we now have 1 IGP, 1 top-level AOP, and about 13 (it doesn't matter) localised AOPs that can choose whether to forge their own path, or promote themselves as subsets of the top-level AOP. The only losers here, really, are the generic CdP producers who now find themselves relegated from the only 'real' Provence AOP to just a generic regional one. This is where we get back to my Russian nesting-doll AVA-approach. Up to them to create sub-local appellations based on microclimates and terroirs, and follow the path of the former DQCs. Perhaps an 'Hauteurs de Provence' sub-AOC for northerly vineyards over 400m that have less ripeness and fresher acidity? (side note: I use this example because in practice the current DQCs have, in my mind, very little typicity or anything setting them apart, whereas elevated northern Provence is actually surprisingly different, much to the chagrin of the winemakers there, who all too often fail to appreciate what a boon this is and instead expend ludicrous effort quashing their terroir)
I've talked a lot about Provence here, so how would this work out elsewhere? Let's take Bordeaux. It already has a generic vin de pays, IGP Atlantique. It has a top-level appellation, AOP Bordeaux, that most appellations can declassify to. It then has a selection of sub-appellations, such as Medoc and Graves. Why is there no equivalent AOP Libournais, to complement this hierarchy? There's then a whole host of village and local appellations. Bordeaux is a solid step closer to my ideal vision than most other regions. Burgundy, with its top-level Bourgogne AOP, seems promising, until it descends into its mess of villages and climats that I wouldn't touch with a barge-pole.
Champagne and Alsace also complicate things. Both have one key regional appellation, but that is declined in so many styles and with so many local specifications that the appellation itself is more an indication of provenance than of the wine itself, and both lack specific sub-appellations or an IGP. Hervé Lalau) sort of agrees with me that this is silly - the difference between an IGP and an AOP is that one focuses on 'territoire' and the other on 'terroir'.
Maybe in fact there should be no regional AOPs at all, but deregulating everything seems a bit wild-westy and doesn't really serve any purpose other than religious belief in terroir over territoire, especially when terroir can be whittled down to half a field, which obviously shouldn't be the level of an appellation (cough cough Chateau Grillet and most of Burgundy). The thing I like best about a regional appellation is that they are so good at crushing exceptionalism and appellations that don't fit into a wider structure.
On the subject of exceptionalism, the Spanish DOCa. Why create a whole new tier for just one region? I read a very interesting article years ago about the abolition of titles in Austria. The commoner on the street (VdF, IGP, etc) didn't care, because he had nothing to lose. The top dog Hapsburgs (Burgundies, Bordeaux, Barolos and Rioja) didn't care - with or without the title, they were still recognisably Hapsburgs, much as Petrus will sell for stupid prices regardless of AOP (or quality, cough cough). The only losers were the middle and lower nobility, who now had nothing. In the wine world, this is the well-made but decidedly second-tier in terms of fame, like probably a lot of the smaller Loire AOPs. The Austrian story is first and foremost an argument against the total abolition of AOPs, but I'm extrapolating it a bit further: Rioja will be a famous and prestigious appellation, it doesn't need the crown of a sole (or sole but one) DOCa, much like Burgundy and Bordeaux don't need it - they're famous anyway. The only wines that lose out are the bread-and-butter AOPs that provide both quality and value for money - precisely the ones that need the most nurturing and protection.
On a final note, that is much closer to my original point about simplifying appellations, I have mixed feelings about segregating styles of wine into their own appellations. Sparkling Champagne and Coteaux Champennois are separate, as are for instance Beaumes-de-Venise and Muscat de Beaumes-de-Venise. Alsace AOP makes red, white, late-harvest, rosé, what have you, but if it's sparkling shifts it into Crémant d'Alsace AOP. Don't get me started on the fact that Alsace Grand Cru is its own AOP... They're both topics for another day!