Times are changing. What was once conceived of as an orderly transition from IPv4 to IPv6 has bogged down and the current situation is one where APNIC has run out of available IPv4 addresses while the Internet is still 99.7% IPv4 traffic.

As evidenced by Paul's ACM article (which is amusing on many levels), I believe we're now in the phase where the incumbent monopolies and their supporters attempt to actively discourage use of new disruptive entrants. In the telco world in the 80s-90s, this was done with efforts to skew regulation and FUD. I see parallels with the burgeoning IPv4 market(s) and I'm sure it'll get worse before it gets better since incumbents tend to want to protect their monopolies. As in the telco world, I imagine it'll take a few court cases for things to settle out. Unfortunately, that takes time so there's likely to be a bit of chaos before stability returns.

If a private registry is anathema to you, please consider pushing policies in the ARIN region which would facilitate global transfers and reduce the need for a private registry.
Because if the RIRs refuse to recognize rational business transactions, these transactions will be driven either to a private registry, or worse, driven "off the books" entirely.

Since I figure address users don't really care about addressing religion, what really matters is which "whois" databases the ISPs are going to query to ensure the addresses presented to them have the appropriate pedigree. As IPv4 address space becomes more scarce, accuracy of those databases is paramount (and sadly lacking now). My impression in talking with the Denuo/Addrex/Depository folks is that they track down the full chain of ownership before getting involved in any transactions. As a result, I suspect they'll have an advantage in not having to deal with the inherited crap the RIRs have to deal with and their database will likely be more accurate than the RIR databases. Haven't looked at the other "alternative" address markets (e.g., http://www.tradeipv4.com/) to see what they're doing, but would be surprised if they weren't doing something similar. Whether the alternative folks will be able to convince ISPs to use their "whois" databases will be interesting to watch.

However, there is, of course, a wildcard here in the form of BGPSEC+RPKI and ICANN policies relating to who will be considered part of the address allocation tree (I'd consider this part of the regulatory approach at discouraging new entrants)...