did anyone address the main concern brought up in the white paper about the reason we don't have a reverse conversion in the first place?
Yes, many witnesses and stakeholders have discussed it and hardly anyone finds it credible. Literally no one has provided as specific concrete example of how this would work without superhuman market timing skill on the part of the alleged abuser. The 3.5 day price averaging window addresses it. To quote the white paper:
Traders who see a massive run up in price would convert to SBD at the high price (when it is most risky) and then convert back after the correction
This is what requires totally unrealistic and superhuman market timing skill to actually work. The conversion in each direction takes 3.5 days which means the person attempting this strategy would need to not only "see a massive run up", but also know that the price will stay high for most of the 3.5 days, and quickly fall, and then remain lower for the 3.5 days necessary to convert back. If the timing is off, or of the market instead moves in the opposite direction, the strategy will experience large losses. Overall the entire story is implausible. Furthermore, if the strategy described did work, the existing one-way conversion mechanism could be already be used to do the same thing at about half the rate (convert after a spike down, wait for recovery, then sell).
I remain open to the possibility that the may introduce some subtle vehicles for abuse, most likely limited in magnitude, which is why I support including a spread between the conversions (essentially equivalent to a fee charged by the blockchain) as a safety measure, at least initially. Doing so would make most if not all methods of abuse unprofitable (of course, a large enough spread would certainly make all methods unprofitable, at the cost of a sloppier peg, so some reasonable middle ground is appropriate). Given more experience we can narrow or remove the spread later, and to be clear it is not necessarily needed at all (some witnesses do not agree with the need).
Clearly the original design in the white paper does not work correctly, even by comparison to other blockchain-based stable coin designs (which have generally worked quite a bit better), or we would not be having this conversation with SBD sitting at $6 or higher for months. The white paper got this wrong (just as it got other things on Steem wrong which have since been changed). It is not a holy book.