So I had a bit of trouble with some inconsistencies in the subwave count of wave A, and being a noob, this is the best thing I came up with (I may be very wrong, if so, please correct me).
The good news is that this different count makes sense and that also it still leaves the possibility of the original A_B_C correction as we expected, with C being under $3000.
The green vertical line is the date that as far as I understood was what Sheba Jafari said the correction should last at the least to (very iffy, might be wrong to either side).
So with the problems with the wave counts - afaik there is no possible 3-5-smth correction. And our A wave is clearly a 3 wave (unless the "news-shockwave" pattern invalidates the waves count), and this B wave has 3 problems: 1) wave 4 is very very close to the wave 1, which is undesirable, 2) the time it takes to make waves 1-3 = time it takes from end of wave 3 to end of wave 5, which is not respected in this case 3) wave 3 and 4 are very small compared to massive waves 1 and 2, which is also making it kind of questionable.
So, if we think of this B wave as a 3 wave, we get a bunch of stuff resolved, particularly - we avoid all those small nuances that I just stated above that make this a questionable impulse wave, next we avoid having a 3-5-x correction, which makes no sense, we uphold the major A-B-C correction.
In case this correction happens to be a zigzag type, we will see C wave going lower than A wave end and the most likely targets are indicated on the graph. 2400s being my most likely C end, but 1800s are also on the bench.
In case we get some bullshit type of correction as a flat (which so far it looks to be), C will still be lower than A, but just slightly, thus making the target around 2800s.
In case we get some triangle correction, I probably won't be able to figure out the end of wave C, but then it wouldn't have to be lower than A wave.
Good luck!