i would take the t-motor, should have slightly more torque

Hi mate, yeah I like the finx23 motors, but even though they are massively light at 24grams per motor, hey are still nearly 10grams each heavier than 1407 and will need a bigger lipo with more weight, I suspect the performance is good but was trying to stick below 250gram here.
Answers to Qs:

1. You mean a “+”frame? Never flown one tbh
2. Oh my, ok one of my pet rants, people go on about COG, they should be talking about COR. Your flight controller checks the gyro for uncommanded movement what? 8000times a second? Up to 32000 times nowerdays, and then corrects it, battery above or below is not the factor people make it out to be. Where it is a factor is:
   Form factor - if it can make the quad smaller or lighter then that is a plus
   Damage - we all have lipos with screw head dents in them
   And most importantly the pilots confidence - talk to Top sportsmen they will tell you anywhere between 20 and 80% of your performance is your mental state, if you are more comfortable with the lipo superglued to the back of the quad then you will fly better with that
3. The proton as listed is smoother (blheli32 with 48khz pwm) about as fast but less agile (although it carries less momentum and has more agility than most 5 inch, it still is like flying a bus compared to this. However I have changed my proton, it now has the 5 inch race arms and brother returner r5s, 5is makes it a more conventional light racer, it has increased the weight slightly, and has dramatically reduced the flight time (because I run more aggressive props) but has more speed, even than this I would think.

Be interested in you tying cyclone 5040s on your light build,

Sorry, I was about "stretch-X" just edited initial question.
thank you for answers!

Lol ok another pet peeve, stretched x - for years when a frame wasn’t perfectly symmetrical we put custom mixes into the flight controller to make it fly like it was. You can use this to make a pure x fly like a stretched x and vica Verica, the actually geometry of the frame become irrelevent
The suggestion some make that stretched x gives clean air to the back motors is daft, we don’t get dirty air from the front into the back motors, not on any quads, and particularly something like this which flew almost vertically.
So once you know how to use custom mixes, the frame geometry has the same impacts as the lipo position-
-form factor/weight and drag
-robustness (not sure which is more robust, Suspect it depends how you crash
-psychological impact - the most important one

-- battery location:
I agree that FC can handle any CoG, but impact is efficiency.

-- stretch-X vs true-X frame:
I completely agree about clean/dirty air: it doesn't matter when quad is vertical :).
But at least you need less tilt when your quad is at lower speed. It's always a balance, I'd like to compare..

Be interested in you trying cyclone 5040s on your light build <-- just received Cyclone Props T5046C, will share the results.

Tilt is a product of thrust vs weight mate assuming you want to maintain altitude, lighter quads need more angle for the same thrust
5046 are good props but are gonna be lipo hungry, the new t5040c are blindingly good, for a conventional 5 inch they are imho better than hq5040 for acro, but on an ultralight they give you damn good speed, I have flown them on finx 23, absolutley worth trying a set, would expect a slight reduction in top end but massive efficiency gain, allowing for longer flight or lighter lipos, the latter would of course increase agility/speed

What do you think about my mock: https://rotorbuilds.com/build/10851
already ordered carbon cut.. I'm interested to compare side by side..

hard to tell from the mock up how mach carbon is in the middle, but concenrs would be too much (adding weight) or two littoe allowing twisting or breakage, but as i say cant tell from the mock up

agree, main plate is 4mm + additional bottom plate 1.5mm. Let's see the final frame.

It's a fair question, and I will have to give it more crashes/time to give a definitive answer. However, it is 5mm carbon, it has less flex than racers with 3mm carbon plates, it is under 250grams AUW and the forces in a crash are directly proportional to the mass, your average 5 inch goes for about 400grams to 600, this being half that, the forces involved when it crashes are halved. My previous sub 259gram 5 inch has the same width of arm. I have crashed that a fair amount, without counting the damaged props, it has twice wrecked motors in crashes and the frame, other than scuffs on the carbon, has survived completely intact

That's impressive! I'm continually shocked at how well my light weight crafts take crashes. Like you said, force is proportional to mass. I didn't realize that was 5mm carbon.

Hi mate, rotorgeeks kindly sent me both to test (3200 and 2800kv). It was very prop dependent, on both 5030 and 5045 the 3200 were slightly faster although there was a bit of a plateau at the top end on 3200 with 5045, 5040 were the fastest props but the 3200kv at WOT were again plateauing and kicking the hell out the lipo, with dramatic voltage sags. On 3s you get a fast agile quad with good flight times with either so would probably run 3200 for that extra bit of speed, if you want the fastest though with viability for racing on 4s then 2800kv with 5040 props is the sweet spot.
The best quads I have (and I have a lot) are my kinetic aerial air-ro running F20ii, they match most of my 5 inch for speed but have the agility of micro due to the low weight. Thirs sort of 5 inch have the same agility but more speed. Hence for freestyle/acro an ultralight 4 inch with F20ii is, imho, pretty unbeatable, but this has an advantage for racing (assuming the track allows you to use the speed)