Also, then there's also the massive size of the transactions. Is there any possible way to get the same security with smaller sized transactions?
Unfortunately no, not possible.
It's painful to see a 20kb transaction, as one single transaction, go out to thousands of people. The size of these transactions far surpass moore's law by at least an order of magnitude. So, is it possible to get reasonably sized, high mixin, transactions for 2kb? I'm talking a mixin of no less than ten with an average of 2-3 inputs?
Moore's law is a worthy discussion -
the average Bitcoin transaction is 250 bytes, and Moore's law says that "it" doubles every 2 years (Moore's law is too slow to apply to everything, mind you, Kryder's law is typically applied to disk storage density, and that predicts 40tb drives costing $40 by the time we hit 2020). If we go solely based on Moore's law, though, we get January 2009 (Bitcoin genesis block) to today as around 6 years, so a transaction size growth of 3x = average Monero transaction size of 750 bytes.
Here's a similar, recent Bitcoin transactions:
https://blockchain.info/tx/239a93ab087b57b6e4ab98a28030546e88c976bd5d66d2eae2bf311ada9d98b8That transaction has 122 inputs and is 18.589kb. To give you some baseline, the 22.783kb Monero transaction has 12 inputs, but with a mixin of 28 that's a total of 336 signatures. It does seem that like-for-like Monero has smaller transactions -
this Monero transaction with 5 inputs, no mixins, has a size of 763 bytes,
this Bitcoin transaction with 5 inputs has a size of 816 bytes.
The tough part here, is that for cryptonote to work as intended, we're outpacing moore's law by at least two entire powers of ten. For cryptonote to even work at all, with minimal anonymity, we're outpacing it by a single power of ten.
This is a tough one - which is more applicable? Kryder's law or Moore's? What about bandwidth?
The bottom line is this: Bitcoin's transaction size is only going to increase if privacy becomes an issue and people start using mixers or similar as a daily occurrence. Transactional privacy requires a trade-off. It's clear that ZeroCoin's trade-off is too large *right now*, but in 10 years time maybe even that will be acceptable. For the moment Monero's transaction size/privacy trade-off is at an acceptable ratio, and whilst we aren't blind to alternatives that may allow us to reduce the per-tx size we're also not going to chase a linear reduction. Our efforts at reducing size implication would be far better spent on finding a solution, in the future, to completely pruning the blockchain.