that it's difficult to know where to begin - so I'll just address them in the order in which they appear:

1. NetApp's 'excuse' was a bit more specific than you suggest: it was Sun's demand for $36.5 million from NetApp to license patents that NetApp denied it was infringing, combined with Sun's refusal to consider the alternative of a licensing exchange based on the patents that NetApp believes that ZFS is infringing. Asking a court to resolve the matter rather than allowing Sun to continue brandish this threat was entirely reasonable and the act of a firm confident of its legal position

2. ZFS offers only modest, rather than compelling, advantages over using existing open-source file systems "to replace expensive, proprietary data storage systems" - and in several identifiable areas (small parallel accesses to parity RAID, for example) pays a performance price in return. Industrial-strength file systems from IBM (JFS) and SGI (XFS) have for many years now supplemented the respected ext2/3fs Linux file systems in providing very credible open source file storage platforms: suggesting that ZFS's respectable but in truth marginal new features significantly change the landscape in this area rather than modestly extend its reach is poppycock. Open source solutions are largely limited to low-end file servers primarily because providing turnkey high-end file service requires a lot of specialized support that most corporations find it well worth paying for (and, again, while ZFS offers a bit of help here it addresses only a small part of the total).

3. Considering that Sun refused to consider NetApp's proposal to cross-license, flat-out denied that ZFS infringed upon NetApp's patents, and demanded $36.5 million from NetApp (see http://www.netapp.com/go/Sun%20Lawyer%20Email.pdf, and read it *very* carefully since the 'demand' is couched quite innocently if you manage to ignore the use of the key word 'defense' - which makes it clear that the writer understands that some kind of defense from NetApp will likely be required), Schwartz's contention that "Sun attempted to avoid litigation with NetApp over ZFS" rings rather hollow.

4. You appear seriously confused in your contention that "once the open source cat's out of the bag, it's out." The issue under discussion here is not secrets but patents (which in a way are the antithesis of secrets: in order to acquire patent protection, you must make the details of the invention public). If NetApp's patents are applicable and enforceable it won't matter how widely Sun has distributed ZFS or under what license: NetApp can legally prohibit its use (and the use of the protected technology) unless users meet reasonable NetApp licensing terms.

One certainly must give Schwartz credit for one thing: he's managed to manipulate a very vocal (though it's not clear exactly how significant) portion of the open source community into taking a strong position in what is in fact a relatively mundane inter-corporation patent squabble. It's usually the case that True Believers make the best patsies, and all a canny corporate honcho has to do is find a way to create an "us vs. them" perception to gain an unpaid, strident group of supporters. Like Mulder, they *want* to believe and vigorously resist any information that might interfere with that desire - so a number of supposedly technically-competent people have jumped onto the ZFS hype bandwagon that Sun has been energetically pushing, especially given the open-source halo that Sun is attempting to endow ZFS with despite the difficulties in reconciling its license with the GPL.

Of course, not all open-source advocates are that gullible - see, for example, Linus's view of several aspects of this matter: http://lkml.org/lkml/2007/6/12/232