Personally I find it very unlikely that we will see the 600wh become 800wh and 800wh to 1kwh for the same size and weight. What this means is we would need to see roughly a 25% jump in cell density for the 800wh pack and 33% for the 600wh. Also worth noting the density % should match for both, if the 800wh sees a 25% jump so should the 600wh.
Right now we know DJI uses the LG M58, rated at 20.88wh nominal, 20wh minimum (looks like DJI claims the min spec for their batteries). Using the 800wh pack with it's more believable 25% jump they would need to use roughly a 7ah 21700. While companies have mentioned a possible cell in their future lineup year(s) down the road nothing has even been spotted in sampling testing, much less full production to the public yet.
What we do have though is 6.5ah 23.4wh cells, FEB has just started sending out samples, and Molicel also has a cell expected Q1. Both are very very new, leading to my skepticism another .5ah will happen within months. I find it way more likely these cells or ones around that capacity will be used in any upcoming battery release. What this would mean is the 600wh would become a 702wh (690wh min claimed) and the 800wh becoming 936wh (920wh min claimed). While retaining the same battery size and a very small weight gain, 30g / 40g.
Because we know DJI runs a 36v 10s system we can pretty much say for sure a 1000wh battery isn't going to happen without a huge weight and size penalty due to the current 40 21700 cells max design. If they did want to they would need another parallel group for 50 total cells. And assuming it could fit would bring the weight estimated roughly to 4.65kg, 1kwh claimed for the current cells, 1150wh claimed for M65A's. Unless DJI goes up in voltage, but I find this unlikely for the time being especially if they want to maintain compatibility with the M1.