Interesting idea. I will skip pointing out the incorrect points in your article, I suspect others will gladly do so, instead I’d like to concentrate on evolving the idea with a couple of random comments:
- There are much less SPV nodes and wallets than you would think, wallets, like Electrum or Mycelium are moving (moved) towards centralized, hybrid SPV models, with a couple of exceptions, like MultiBit. This is not an argument against what you say, since there are still plenty of SPV nodes out there, just something to keep in mind.
- On privacy: if you are not using a full node today ALL your addresses are already tied together by various third parties. Blockchain survelliance companies are sitting on their data and it’s just a matter of time when the techniques they use become available to the wide public. This is a timebomb. The reason is because a full node gets all the data the network have, therefore network analysis cannot be applied here. There are only 2 ways to solve this without running a full node (I reviewed them all). One of it is my idea, which requires centralized trust in a public full node on the legitimacy of a transaction, but it still wouldn’t be able to breach privacy. My initial guess it could provide full node level privacy for mobiles before experiencing serious syncing times up to 10MB block size. If you add SegWit, Schnorr and other improvements, doulbe/triple the block size with additional benefits the maximum block size you are talking about could be around 3MB without issues. If it’d be me I’d like full node level privacy implemented with my idea or the only other idea that MIGHT help, called committed bloom filters.
- Making SPV wallets enforce rules seems to me a really hard job. We’d have to convince SPV wallet companies to actually implement it, dig into their legacy codebase and implement rules those are probably haven’t ever been implemented in the programming language they are written in. Also the resistance would be huge, since wallet providers don’t care much about network health, but rather the convenience of their customers. This would end up in millions of dollars in costs.