Marc, I think you are correct. Anonymity is not a big market (and I am formerly @AnonyMint). But do realize all our crypto markets are tiny right now, so Monero could still have upside.
I think you all got that idea about "anonymity" backwards. "anonymous" crypto is not about secret agents, terrorists and shady markets (although they do appreciate it of course). People don't seem to realise how much LOSS of privacy open block chains imply, as compared to more traditional ways of doing things. I illustrated this in several posts elsewhere, but bitcoin-like block chains *propagate* too much partial information. I gave the example of me, the plumber, repairing the toilets at Joe's bar, getting paid in bitcoin, and then offering those coins to Mary, my friend, who then gets a coffee with it at Joe's bar. Joe (and several of his customers) can now easily find out that Mary is my friend and that she got the coins from me. In NO OTHER PAYMENT SYSTEM such a lack of privacy is present. If you had paid me with fiat (whether cash or a wire transfer), he wouldn't be able to find out that I gave exactly his money to Mary.
What "anonymous" crypto does, is putting back some privacy into the block chain. People have been attacking coins like monero because of their inability to be totally untraceable when the NSA is after you. But that's not the point. The point is that the so-called pseudonymity of bitcoin fails when one can link all transactions, combinations of change addresses and new payments in one big, cluttered network that propagates all partial knowledge about individual transactions. Anonymous coins are not about being able to pay evil deeds without being annoyed. They are just about not being a tool that leaks partial transaction information all over the place.
The big market is scalability and some use case that would propel crypto-currency into the mainstream.
I think that the big market is finding a use case. Clearly, apart from betting and speculating, crypto doesn't go anywhere. The hypothetical case of me, the plumber, being paid in bitcoin for repairing toilets in a bar is, at this moment, still a fairy tale. That's also why it SEEMS that anonymity is not a market. Nobody cares, because bitcoin is, at this point, like most crypto, just a kind of backing of exchange IOU.
Note as transaction fees approach the minted rewards, PoW loses incentives compatibility and consensus diverges. As block sizes increase security decreases because propagation becomes a larger percentage of the block period.
This is only correct if the increase in block size goes faster than the increase in network bandwidth. If block size doesn't increase faster than Moore's law, there's no problem, and if the protocol is scalable (doesn't contain hard constants), then the time taken to transfer them remains constant and small compared to the time spend on PoW. Also, if blocks get larger, the corresponding transaction fees get larger too. So minting a large block with a longer time can be just as lucrative as minting a small block.
In fact, the amount by which you chop up the chain in blocks doesn't matter. Whether you process 100 blocks of 10 MB in 10 minutes, or you process 1 block of 1 GB in 10 minutes, has about the same network load, the same ratio of network time over PoW time. The only difference is that with 100 blocks in 10 minutes, you get much faster a confirmation of your transaction. Note that the PoW that "safeguards" your transaction is also independent on the block size/time ; only the total amount of PoW spent after your transaction secures your transaction, whether that's chopped up in 100 small efforts, or one big one, is indifferent. Of course, 1 GB of transactions in 10 minutes is only, as you rightly point out, reasonable if we can blitter 1 GB to China, to Brazil, and back in, say, less than a minute. Which will happen, according to Moore's law, at a certain point in the near future.
In fact, I think that continuous forking of important coins into several competitive ones can also solve the scaling issue. Look at ETH. Not that they have a scaling issue at the moment, but the ETH network just doubled when it split into ETH and ETC. If such kind of forking would happen continuously (say, on average once a year), we would get many competitive coins with similar user basis, and be able to distribute the load over several independent networks. Exchange points (preferentially decentralized but even if they are centralized that's not a problem) would get us from one chain to another ; like we now have different fiat currencies. Market forces would put prices on the different networks. Note that for "hodlers" that would be a night mare, but for use as a currency, that's no problem. It would also create enough inflationary pressure to keep the price of coins down, and stop the silly speculation.