I haven't looked into it, but why wouldn't the proposal account also earn interest?

I'm not sure, but if I remember right, it only pays interest on SBDs in "Savings". I'm assuming that the SBDs in the SPS are not held in savings.

I think ludicrous interest rates would have a tendency to make the chain look bad to outsiders

Perhaps, but I'm not sure if it would look more sketchy than our Trending page already does. And it doesn't need to be permanent. The main ideas that I wanted to explore here were: (i) the possibility of using interest rates to peel apart the interests of delegation-bot clients from the interests of the delegation-bot operators; and (ii) that I think the interest rate doesn't apply to all SBDs, just to the ones that are held in savings (which could still be wrong). But yeah, there are risks and tradeoffs to be managed.

It's not at all clear to me that the bot businesses would flow their assets from SP to SBDs, why not just maneuver the price of Steem to guarantee SBD printing and use their SP to make sure they get a big chunk of the even-more-valuable-now SBD rewards?

If they did that, at least it would be good for the price of STEEM, even if the content quality stayed the same. In general, though, I suspect that most delegation-bot clients would prefer to have entirely passive rewards.

Also, the more valuable SBDs are the more tempting it could be for someone to raid the SPS funds, and if you're right about big holders powering down to get passive interest then more cheap liquid Steem would be floating around for someone to buy up to start controlling proposal (and witness) votes.

Those are both good points that I hadn't considered, although "cheap liquid Steem" here is not consistent with "why not just maneuver the price of Steem to guarantee SBD printing". The problem of "cheap liquid STEEM" could be managed, perhaps, by increasing the rate gradually in order to spread the powerdowns out over a longer period of time.

You may also want to think about any exchanges that list SBD, they're probably large SBD holders from the chain's perspective.

Also a good point that I hadn't considered. I'm sure you're right.

In the end, there's a big range between 0 and 100. I'd just like to see witnesses considering the full range of possibilities. Maybe I'm wrong, but it doesn't feel (to me) like anyone has seriously thought about this parameter in quite some time. We have a lot of experience to suggest that 0% doesn't do much to stimulate growth, but we don't know much about what happens at other values. Even the Hive example doesn't directly compare because they've changed many of their other rules, too.