I've just tested my lightest flight arrow yet 375 grains, it's 5/16" shaft (30-35 spine) 28" long tapered at each end with a tiny brass point. I tried it from the 70 pound Yew longbow bow, it's difficult to see exactly what it does over 10 yards, but it seemed to leave the bow and snake violently coming back straight just in time to hit the boss.
I think it would be hard to get the weigh down much from that as if the point is much lighter it would make the centre of mass too close to the geometric centre. Oddly the Turkish flight arrows had the centre of mass behind the centre of gravity! However work by Clarence N Hickman concludes it should be forward of centre. Mind he wasn't testing arrows the same length or shape as the Turkish flight arrows.
For the record a bare 5/16" shaft (32" long) from the same batch weighs 280 grains, which give some idea how light the point is.
If I get time tomorrow I'll see if I can get some footage with my camera which will show it flexing... dunno if it will work.
The bow seemed happy enough with that weight, it didn't really jar or ring. The acceleration is pretty huge so even a light mass has enough inertia to stop it being a dry loose. I did some rough arithmetic a while back and the g force on the arrow was surprisingly high.
I just re-did my 'back of the envelope' calculation assuming a constant acceleration* from 0 to 180 fps over a 21" power stroke I came up with an acceleration of 289g !!!
This seemed mad, so to check it I enlisted the help of Mr Google and came up with this article which shows some practical tests using a lower poundage bow and an accelerometer which gave results of about 200g.
Note in their tests the accelerometer has added considerable mass to the arrow which could account (along with the lower poundage) for the discrepancy between my figure and theirs.
So, under 200g acceleration the arrow effectively "weighs" 10.7 pounds so it's nowhere near a dry loose.
Now these figures probably sound bonkers to you so please feel free to check them and correct me if I've slipped up.
*Note:- Acceleration from a longbow is not linear, it is highest over the first part of the power stroke where the draw weight is higher. This is why Clarence N Hickman designed the modern recurve bow (with the limbs angled a little towards the archer and then curving away in deflex reflex) he wanted to get a more even, linear acceleration over the whole power stroke rather than the harsh kick up the backside that you get from a longbow. He was after efficiency so that he could get decent arrow speed from a lower draw weight. Why? Because he'd lost some fingers in a rocketry accident, he's the guy who designed the Bazooka and did much work on rockets in WWII.
Of course the cams on a compound allow the designer to manipulate the acceleration curve.