Surely the SPV mining pools will want to do that, then, if it works. If it is in their interest to do so...
Sorry if you felt that you'd found some magical anarchist utopia without rules, you're wrong.
Those rules are reasonably well known - and bitcoin core is an example of where to find them - you should try learning about them one day.
e.g. I can't go mine a zillion, 1,000 difficulty blocks and get 10's of thousands of BTC a day with a 1THs miner.
Damn that's not what you thought is it? Yet that would be more profitable if I did that wouldn't it?
Also, of course, there's the block header requirements, like versions and hash values based on previous blocks and merkle tree hashes of transactions.
... and for those transactions ...
Yep, again, wouldn't it be more profitable to make a transaction that created 1,000 BTC and sent it to my address any time I wanted to?
So yeah, anyone can do those things, but then they are no longer mining Bitcoins and (most? ) Bitcoin exchanges wont accept their scam-coins
SPV mining allows for their blocks to fall directly into this category.
As I already stated, they don't verify the transactions and merkle tree and ... ...
So if someone did indeed make a block, with a transaction in it that created 1,000 BTC and sent it to their address, and pushed out that block to the SPV pools, guess what? The SPV pools would accept that block and start mining on it.
Yep their method will lead to breaking the rules ... and they have already done so in the past.
I insist, it was not their fault, but the fault of the devs who programmed BIP66 to be enabled immediately when 95% of the miners had updated, with no grace period -- thus ensuring that 5% of the hashpower would still be using the wrong rules at the fork time, no matter how quickly they tried to update.
That 5% of the hashpower would have been wasted anyway, even if the SPV miners had not mined empty blocks on top of that bad one.
Bitcoin is not set in stone to never change for eternity.
It will, and must, change over time to accommodate the requirements of the majority of the bitcoin network.
With Bitcoin, it has been done via a trigger in the core code that makes the change effective when the requirements exceed 95% of blocks mined over a given history of blocks. That's how it was done for the v3 change. The grace period was the whole period leading up to that 95% acceptance.
A longer grace period would have made no difference to their SPV code.
Their code didn't bother to even check the block version number, so whenever in the future, after the v3 change over, if someone mined a v2 block and sent it to the SPV pools, they could have accepted it and forked the network. Yes it was their fault. Yes, they had updated bitcoin daemons. No their SPV code was wrong.