Nothing official from Luminus.
That seems normal, because Luminus often releases some products to market for beta testing before putting them on the official site.
Most likely just a slightly more powerful version with just a better flux bin. Marketing!
There usually has some physical changes between each revision and it doesn’t look physically any different to the original.
Also the very high VF of SFT-25r doesn’t inspire confidence. What if the new SFT-40 also has a very high VF? That would be bad.
I think for single cell throwers, a 6V throwy emitter with smaller LES is what the market needs. Otherwise we will always be limited by the anxiety of Vf.
That is pretty much impossible to make, however. You will either have 4 tiny LES, where the gap between them leads to ugly beam patterns, or stacked dies like 719A, with the same issues (heat sinking and low efficiency) of 719A.
Indeed, didn’t think about the gap issue.
If the chips are flip chip design it could be possible to make a very small LES without huge gaps in it (or at least fill the gaps with some special phosphor-silicon mix to make the gap much smaller). I don’t know of any LED chip that is extremely small and at the same time has a flip-chip design.
But anyway, I think such a hypothetical LED would be very difficult to manufacture (especially in bigger scale) and therefore expensive, which no one would buy in bigger quantities.
Might it be possible to address the center gap issue with a 3x3 design?
As 2s requires an even number, this would lead to uneven illumination with 4 LES& 5 LES.
But one could reduce the surface of the four corner LES by a quarter each, and combine them with the center LES, for a 4 LES & 1+4*0.75 LES setup.
Is this possible from a manufacturing perspective?
I think so. Since two different chips are needed (which could be different in electrical specs), this hypothetical LED would be expensive. Also I am not sure if the established production processes could be used to manufacture such an exotic LED in high quantities.
Is the Sanan 43mil mentioned in this post flip-chip? If 43mil is the side length of a square, then if there are 4 dies like this, it is basically equivalent to a SFT-40 but a 6V version.
I don’t know much about LED design, it just occurred to me.
The Sanan LEDs that I have appear to be traditional construction, and not flipchip
I don’t have the SFQ43 specifically, but I have the SFN43.
Yes, the 43mil is the side length of the LED chip and its a filp-chip stuff, I did not see any latest generation chip with traditional vertical process and require bonding wire at fronts.
Also, San An’s 43mil chip have several generations and the early one is totally crap with higher Vf and much lower efficiency.(For example, a LED constructs with 16*43mil chip[Early SFH43] will only give you 6000lm at 120W)So my LED uses the latest one to achieve best performance.
Wtf 
A 2x2 with these latest gen chips would be interesting! Would be ~ the same LES size as a SFT40 (4mm²) but in 6V or 12V option…
Yes, especially in a 5050 or 7070 footprint!
I’m pretty sure the Osram W1/W2 are flip chip, it doesn’t look like there’s space for bond wires in the package
No flip chip. This is classic lateral design. (Here CSLPM1.TG shown)
You can see the missing bonding wires on the left side of the LED chip. These wires are covered with the white epoxy resin so they are not visible without disassembling the LED.
This LED chip and overall design is very similar to the LE UW Q8WP I tested several years ago.
Nichia 149A uses a 9 chip design and 9Vf, but I’ve never seen it used, it’s relatively old now and the footprint is weird.
https://led-ld.nichia.co.jp/en/product/led_product_data.html?type=NV9W149AM&kbn=0
Very unique. Power-wise it looks like a perfect match for the D3AA driver! This could be a great match for a single LED version of the D3AA using a larger reflector or TIR.