Sorry for username change, email adress issue.
I'm wondering about coverage, because given the very precise needs of the problem, a bearing being covered by your antenna beamforming abilities might not be enough, you need to be able to match an amplitude, phase and frequency at that bearing. Distance between phased array elements is directly related to the frequency range it can emit in, can the SPECTRA arrays meet the frequency agility of modern fire control radar ? I'm not implying it's any kind of engineering oversight, just that fighters are inherently limited in their jamming abilities because of power and antenna size.
Active stealth has huge accuracy requirements on the estimation of the incoming signal, and there are hard limits on interferometry based on array design (azimuth resolution is related to gain, which is in turn related to array size). AFAIK, SPECTRA's azimuth resolution is about .5 degrees, which is excellent, but that's about 1.5km uncertainty at 30km distance : Is it enough to meet the accuracy requirements of active cancellation ? You need to minimize amplitude error, frequency error, and phase error to accurately cancel the incoming signal, to exactly reproduce the extremely complex incoming signal, then phase-shift it and hope that your antennas allow (gain-wise / frequency-range-wise) you to match the frequency / phase / amplitude everywhere around the plane, without producing any unwanted interferences that would increase the radar return instead of decreasing it.
Modern radars use LPI techniques : modern radars will use frequency-hopping, mode-hopping (simultaneous SAR / air search) etc. To do active cancellation, you will need an extremely precise model of the emitting threat, by nature a very moving threat, especially with modern, digital radars that are intended to easily integrate new signal processing software. Is this threat library available in all relevant cases ? Is it even possible to keep up to date ? You are not always going to get an intelligence edge on your adversary's radar capabilities, and what is your active cancellation going to look like if suddenly, a software update changed the way your target radar hops around frequency wise ? It also raises the export question, because threat libraries are usually not exported AFAIK. So you need to model the incoming signal *very* accurately, but the emitting device is by nature shrouded in secrecy, which makes it even more difficult.
ultimately, this message is not a statement on the skill of french EW engineers : I'm not thinking that Thales couldn't do it because they aren't competent, I have big doubts on the technical feasability of the solution on any plane. The fact that most countries with an ability to design aircraft seem to take some form of passive stealth as a given (NGAD, SCAF, J20, Tempest for the relatively recent and future projects), even with the huge engineering tradeoffs associated (you pay in aerodynamics for reduced RCS), points to the fact that it seems to be the direction everybody is taking for evading radars.
with regard to other manufacturers marketing it : ultimately, the wording is very vague (https://www.baesystems.com/en-fi/article/stealth-master ) : "digitally hide its signature, becoming invisible to radar, or to digitally create a complex and confusing picture (noise) for a threat operator" which is basically a broad definition of active EW. For the Gripen, even with the hype around it's EW suite, SAAB did not dare to say that it implemented any form of active cancellation (at least I did not find any, I'm of course very open to being proved wrong).