Yeah, but why not do it with their own funds if this is what they want?
Because it has a cost that results in a benefit to Steem and not to a privately owned business like Steem Monsters.
Since the benefit is to the platform, it makes perfect sense to use platform/community funds (aka reward pool and/or SPS) to pay for it.
And why their argument is that buyin it in one swoop won't work, but their way of buying slowly will?
The argument is not about buying it slowly vs buying it quickly, it is buying it repeatedly vs. once.
65k STEEM is not enough to increase SBD to $1 and keep it there. In order to get it to $1 will require a lot more than 65k STEEM worth of SBD to be bought and destroyed (already about 50000 STEEM worth has been bought and the task does not appear to be close to complete). The faster you try to do this, the more it costs due to slippage (you pay more STEEM for the SBD than you get back from converting it), so less STEEM is available on the next cycle and the process slows down overall (less SBD is bought and removed from circulation over time).
RE: The @sbdpotato beneficiary experiment