Hi Don,
I fully respect your work but Larry or you, already said that you have version(s) with more than 20 ELO improvement over Komodo 4.0
I think when commercial interests interfere with the pursuit of progress at true pace, this is when the chess community at large starts to suffer as every author is locked (or shall I say thread deadlocked) to everybody's else first move.
As for the political answer from Robert, I don't buy it either. Of course burning DVDs, advertising and distribution takes time but, hey, we live in a global internet village. Updates, as you do, can just be uploaded and fetched from a website (with possible mirrors). People can have a base version shipped on DVD and upgrade to the latest subversion as soon as one is ready. As for billing, each subversion could be priced to say $9.95 over the base (yearly) version for a 20Elo increase update or half a buck per ELO point!
The real burden is on rating list maintainers as more releases will make their testing more demanding. It will make watching the lists much more funny too as for months there were no remarkable change at the top.
On the gentleman agreement proposal, I think it will benefit everyone, provided it is genuinely observed. Larry rightly pointed that part of the problem for accurate progress assessment was the availability (or lack of) of strongest engines to test against, hence my post.
I'm just proposing a scheme to climb the ladder with "easier" steps than the effort mandatory for "giant" (50 ELO) strides