Especially since the advent of data-linked passive radars is expected to render shaped stealth irrelevant by 2030. That is not a conversation for here though!
While the subject is somewhat OT for the RNZAF thread, since it was brought up here I will do a brief rebuttal along with links to old DT threads where discussions were held on the topic.
Firstly, the terms passive radar and stealth are buzzwords which tend to get used by sales/PR people for defence systems, or those who do not really understand the concepts they are being used to describe. As such, when I see someone using them, or making the sort of assertion which was made, it tends to raise a warning flag with me.
To describe something as a 'passive radar' is false, since a radar cannot be passive, a transceiver has to actively send an RF signal which then comes in contact with an object, and the transceiver then detects the reflecting radar 'return'. What is usually meant when someone uses the term 'passive radar' is either a (passive) ESM which detects RF emissions from other sources, or (an active) bi-static radar array, where the transmitter(s) and receiver(s) are not integrated into a single unit, or co-located.
Low Observable or LO is the correct term to use, not stealth.
Now there is the potential for large area integrated ESM and/or bi-static radar arrays to be used to detect LO targets when coupled with advanced datalinks and computing power. In fact there has been some anecdotal evidence that one such radar array (JORN) has managed to due just that. However, such systems cannot do much more than provide a sort of tripwire detection capability at present, and given how such detection is managed, this does not appear likely to change any time soon. Such systems cannot realistically provide track data for a target, never mind targeting data, so such systems could not guide a datalinked SAM to the target. Until a sensor system exists which can reliably detect, track and lock onto a LO target for the entire duration of a missile launch and flight, while providing datalinked in-flight information to the missile so that missile never needs to rely upon it's onboard seeker(s), then design elements (including shaping) to achieve LO are still going to be very much relevant.
In terms of the expected sensor improvements over the next dozen years... such a claim is illogical. For one thing, it ignores the fact that papers, studies, and some testing have been done since at least as early as 2000 which showed the potential capabilities of such systems, as well as their very real limitations and exploitable weaknesses. For another, if 2030 was the expected time frame for a significant portion of current work to achieve LO to become obsolete, then why would the US expect to be fielding some of the current LO designs out until 2050? Similarly, why would a number of other nations with active LO programmes for either manned or unmanned aircraft still be working on systems to bring them into service, if they are going to have a less than a dozen years service before becoming obsolete. Keep in mind, most (all?) of these systems have not entered full-rate production, never mind reached FOC or had production complete.
Now, for some further reading on some of these topics:
How effective are modern radar systems at detecting low-RCS targets?
Propagation of stealth technology and what this means for the US
Super radar detecting US stealth plane
F-35 Fantasy or Fake F-35 Discussions Debunked
Please note that these are some old threads which have been inactive for years, and some were started over a decade ago. I mention this because I do not wish people to attempt to do a Lazarus on necro-threads. If someone has a question, I would suggest starting a new thread and linking to the old one. Another reason to note the age of these linked threads, is that in most of the threads, speculative or hypothetical systems were trotted out which would 'defeat stealth" and their entry into service was right around the corner... In those cases, the limitations of the hypothetical systems were pointed out and as time would suggest, such systems either have not entered service, or not provided the capabilities claimed.
For those seeking additional information, I would suggest looking for information on Over the Horizon Radar (OHR), with JORN probably being the best known system, or the Sweden's proposed bi-static AASR which IIRC was cancelled in 2000 before a hardware test could be done.
For the RNZAF, and in fact the broader NZDF as a whole, decisions need to be made about NZ policies regarding international cooperation and coordination of efforts. The US is working hard on developing CEC, which as I understand it would involve data exchange between assets on a level beyond what Link 16 is capable of, so that sea, air, and eventually land forces can have a common operating picture. Australia is also interested in this system, so that Australian assets can seamlessly slot into and operate alongside US task forces. NZ needs to consider whether the NZDF wants or needs to be able to cooperate and coordinate to this degree. IMO though, if NZ does not opt to pursue opportunities to slot in, then NZ forces will likely lose opportunities to operate alongside US and Oz forces in the future. OTOH if NZ does decide that such a capability would be good to have, then that would almost certainly impact procurement decisions, since the US is developing CEC for and on their systems, and would likely limit outside access to develop a CEC plug-in, plus the cost to develop the plug-in.