I am what I would consider a Bitcoin fanatic. Beyond the idealistic reasons that I choose to use Bitcoin, I have plenty to lose (financially) if Bitcoin drops in value (and plenty to gain if it increases in value). If I am concerned about my ability (as a Bitcoin fanatic) to run a full node with the current anti-spam (has spam been fixed by the way?) 1MB block size limit in place, what is going to happen when the data I need to share with my peers doubles (or increases 8-fold, wtf!)? When Bitcoin fanatics have doubts about running a full node, I would imagine that the robustness of the decentralized network has been harmed.
Yes, I am of the opinion that I absolutely must run a full node to take full advantage of Bitcoin.
The price of bitcoin has been, and still is, determined by the expectation of its demand being high in some future time.
Until recently, that high future demand was supposed to come from its use as a currency for internet payments. The current level of use (maybe 5 million USD/day, by optimistic estimates) is too low to produce that demand, by orders of magnitude. To justify a price of 2500 USD/BTC (say) one would need 100 to 1000 times as much use.
The current "normal" traffic T, ignoring the recent "stress tests", is about 120'000 tx/day, or about 0.45 MB/block on average; and has doubled in the last 12 months. That is too close to the effective capacity C of the network, currently ~200'000 tx/day or ~0.80 MB/block. (C is not 1.00 MB/block because of the inevitable empty blocks).
Granted, much of that traffic is not payments (bitcoins changing hands in exchange for other goods and services). Lots of it is gambling, wallet housekeeping, tumbling, notarizing, testing, and possibly fake traffic meant to give the illusion of increasing adoption. How much of T is due to those "non-essential" uses? Guesses vary; my own guess is 90% or more.
But even if 90% of T was "non-essential" and could be eliminated, to get to 2500 USD/BTC one would still need 10 to 100 times more traffic than there is today. If running a full node is not viable for home users today, it will be quite impossible by then. In other words, if the proper working of bitcoin depends on home users running full nodes, then bitcoin is not a viable competitor for PayPal or Apple Pay, much less for VISA.
Back to the present, the block size limit would not be such a pressing issue if that "non-essential" traffic were excluded. IMHO, the best way to do that would be to set a significant fee and a significant minimum transaction value -- say, 0.001 BTC (~0.25 USD) per output. That is how Charlie Lee solved the spam problem in Litecoin; but his suggestion to the Bitcoin devs to do the same fell on deaf ears. (So much for the old claim that "Bitcoin will not be superseded by a better altcoin because it can itself incorporate any good features of the latter".)
By my guess, a significant fee and value threshold would reduce T to 1/10 of the present value. It would also make stress tests and spam attacks much more expensive to the attacker, hence much less likely. (We still haven't seen a real spam attack, but it is estimated that an effective one could be sustained for a couple of days for maybe 100'000 USD/day, or less.) Note that 0.25 USD/tx is still less than 10% of the per-transaction cost of mining. On the other hand, it would not prevent the hoped-for 100x or 1000x increase in adoption that would be needed to drive the price up.
Unfortunately, raising the fee is politically impossible, for several reasons. "No fees" has been a pillar of the "marketing" of bitcoin since the drive for general adoption started in 2013, and is still part of the discourse of important players like Coinbase and BitPay. A fee raise would probably break gambling compaies like SatoshiDice, and several other companies that are politically powerful in "bitcoin space" (e.g. by sponsoring the "bitcoin media", events, and avertising in mainstream media.) A large drop in T would also expose the fact that adoption for e-payments is not growing, and probably declining.
A large minimum fee has been opposed also on "ideological" grounds, because it would need some governing body to decide the proper value, and change it from time to time based on the current market price and the state of the economy. Thus, the small-blockians claim that letting the "fee market" define the fees would be an "ideologically pure" solution. But the 1 MB block size limit was not chosen for that purpose. "Market" fees that result from an arbitrary block size limit would be just as arbitrary as fees set directly. A governing body would still be needed to impose a block size limit, and revise it from time to time, to achieve the best fee and volume numbers.
So, what is the solution? In my view, there is none; hence my signature. As most of you know by now, I think that bitcoin should never have been marketed to the general public. It was started a technical experiment to test whether the protocol worked in practice, as it seemed to work in theory; and should have remained such. The protocol mostly works, but the result is not a viable currency for several problems that were not foreseen. Fixing those problems seems to require another couple of brilliant inventions, like the PoW blockchain.
