There is no real tool for downward pegging, other than waiting it out...

The only conceivable method that exists (outside of reward-based initiative such as my @burnpost) would be a positive bias on the price feed, and that is not a good one because it amounts to increasing inflation above the defined paramaters, with not even any future remedy to absord the newly created coins.

I don't necessarily agree with this. The tool for downward pressure when SBDs move off the peg is to reduce or eliminate the incentives to buy/hold (interest rates) and to increase the supply to meet the higher demand (bias). The increase in supply obviously takes time to reach the markets, but it would still provide downward pressure (assuming the parameters are actually changed by enough witnesses to reflect market conditions, which currently doesn't happen).

I think that consensus changes for increased supply is actually a good mechanism and the preferable one if we're talking about token inflation. Even if the new supply ends up creating more than the market demands (which would be the obvious goal for short- to mid-term price spikes anyway, in order to bring prices back down to the desired peg), parameters can then be adjusted to incentivize holding or converting SBDs to STEEM if SBD prices happen to over-correct.

But this is why witnesses should be well-versed in the blockchain protocols and price/token parameters and at least have a basic understanding of currencies/markets/economics. This can't happen if people don't care or don't understand and if they're not paying attention to the markets (or if they're completely absent witnesses altogether). We can have any tool imaginable at our disposal but it won't mean much if people don't know how to or just don't care enough to use them. If the problem is simply that witnesses don't care about pegging or don't know how it works, then none of this even matters and this should be a consideration for our witness votes.

The proposed improvement of adding a method of converting STEEM to SBD does not increase inflation...

Why doesn't it increase? Whether we use bias or use a conversion function, more SBDs will be created. The former at least allows consensus protocols to manage the increase of supply, but both serve the same purpose, which is to meet market demand for SBDs.

I'm not necessarily opposed to a new conversion function. I'd just like to see the current tools actually tried before we condemn them for being useless. In any case, we probably still have quite a bit of time to figure this out and there's no reason why we can't test both or either method - and whatever we try won't be a quick fix. Unless we're assuming that SBD speculators are always up-to-date on the internal mechanics/sentiments/changes of or with Steem, which I believe is rarely ever the case, there's no reason to expect that any immediate changes to protocols or parameters will yield immediate results. I'm OK with a slow return to the peg via slow increase in supply to meet demand, if it can be done.

...and can be reversed later if it becomes necessary...

This is something I'd like to see more of when making protocol changes that yield bad results. But I think I'm dreaming a bit there.