Darkonnis,
I respectfully disagree - as for the whole average AA percentages, you arguably have a point, but the law of averages also suggests that the randomness in the mix in each 100g bag combined with the randomness involved in picking out a portion from the bag means that extreme variations are unlikely. Also, you would to some extent experience trouble in even 23L standard batches as they still are comparitably small batches.
I am not suggesting that you can scale from a 5L trial batch straight to a 1 bbl batch (let alone a 10bbl batch) - if you are brewing at those volumes, I think you would want to fine-tune anyway before making the large batch, if for no other reason to save a bit on grain and hops where possible. However, I do stand by that my 1 gallon batches can be scaled to the standard home-brew length of 5 gallons/23L without major problems - in fact, any variations you might experience would be no different than the differences you might find from one packet of hops to another, or by slight differences in technique from one brewer to another!
As for wastage, I re-vacuum pack my hops between brews using a domestic vacuum packer, and I use said packer to re-seal my bags of malt. Oh, and quite a few recipes at a 23L brew length use less than the full 100g hops, and many recipes use more than one type of hops.
For me the nano-brews work well, as I have limited space, and I can try many recipes and decide what I like, then once I get more space, I can brew them in bigger batches. Ultimately, it's a matter of preference, and as I've mentioned, I wouldn't mind a nice shiny set-up, but I have neither the space nor the money for that! With my small brews I'm still gaining valuable brewing experience and knowledge, and like you said, it's a good way to try out different variables, such as yeast and hopping. The main down-side for me is it's relatively labour-intensive for a relatively small outcome (about 9 450mL bottles!). But it's either that, or nothing at all, so for me it's a no-brainer!
Dennis