Why would lowering the micropayment be riskier than with the current micropayments? (because they would be easily spotted?)
Because the attack would become cheaper, hence worthwhile.
mmmm, in my opinion the attack is already worthwhile, anyway such an attack won't succeed unless the attackee is not vigilant and doesn't double check their address (as the bot instructs them), so why not lower it if it permits to filter out exchanges addresses at the detection process (not requiring you to check suspect addresses)?
People do make mistakes, and having to be overly vigilant scares them off. I'd rather have a more comfortable environment for the entire process.
The attacker's ROI depends on 3 factors:
- % of people who do not check that the address is theirs after receiving a
second notification "Received your payment from ..." (the attacker's microtransaction is almost certainly later than the legitimate one)
- average balance of such users
- cost of attacking, which is the average amount of the microtransaction
By lowering the cost, we turn the ROI at some point from negative to positive.