The magnus effect is very real but calculating it is not quite that straightforwards because it is proportional to the diameter of the cylinder, its length and it's rotational velocity (spin rate) and to the value of the wind and to the movement of the bullet about it's centre of gravity or rotational stability. In other words, any programme or model would need to model 6 degrees of freedom but most apps only model for the X, Y and Z planes and do not include bullet movements on the pitch, roll and yaw axis.
"Aerodynamic jump" caused by the magnus force and the bullet gyroscopic motions cannot be accurately calculated for a number of reasons best left to those who wish to read the Litz Advanced Ballistics in depth explanations further, but they can be approximated as follows: (note, this is taken from Brian Litz Applied Ballistics for Long Range Shooting (2nd Edition) P78 & 79)
Y = (SG/100) -(0.0024L) + (0.032) MOA/W
Where Y = value of vertical deflection, SG is the bullet stability factor, L the bullet length (in calibres, so a 1.5 inch bullet in say 6.5mm becomes 5.86 calibres) and W the wind speed in MPH. (ie full value of sidewind)
This is the approximation for every 1mph of sidewind so for a 10mph sidewind you would end up with the solution multiplied by 10.
For wind from the left as viewed from behind the bullet, this value has a deflection component which is RIGHT and DOWNWARDS
For wind from the right, this value has a deflection component which is to the LEFT and UPWARDS
Now onto practical considerations. Why is it that competition and BR shooters usually ignore this motion? Perhaps it's because they get sighter shots and because this effect being separate from actual vertical wind deflections (ie due to vertical winds as opposed to Magnus effect and aerodynamic jump) makes it almost impossible to deduce anything from grouping patterns due to the number of variables involved.
The actual vertical wind deflections caused by undulations and thermal differences between shooter and target are generally not significant at closer ranges, but become more so when over long distances, there is more chance of undulating ground, and the true value of these true vertical components varies, so understanding their impact becomes more guesswork and experience since they cannot realistically be modelled.
For long range hunters who want to maximise the chances of a first round hit at long range, then it arguably is worth considering aerodynamic jump as well as observing air movement along the flightpath of the bullet to be fired and adding whatever compensations are thought needed. No sighting shots here!
It does raise an interesting point and that is irrespective of how many amazing looking first round hits on coke cans and beer cans at 1000yds plus you may see on You Tube, the chances of a first round hit at those distances owes more to pure luck and guesswork as much as worrying about the external forces and aeodynamic jump by calculation. That should now be more obvious, so the conclusions are to my mind at least:
1. For deerstalking to 400m, it is not worth thinking about and is for all practical purposes immaterial;
2. For competition target shooting, due to sighter shots being available, it is also probably not worth considering, although changeable vertical components of true vertical winds may be;
3. For ultra-long range varmint hunting or for those countries where long range hunting of large animals is practiced, aerodynamic jump should be a consideration for first round shots because it is material in maximising chances of a hit. I refrain from saying "humane first round kill" because there are simply too many variables involved and luck has to play an important part. My personal views are that it isn't something I would do myself. I'll leave that argument at that;
4. For long range load development purposes, considering the tendency of groups to open up in an exponentially increasing manner as distance stretches out, combined with vertical deflection of wind and aerodynamic jump, one can see that it does more to detract from making meaningful conclusions from group patterns and sizes once beyond say 600 yards, than it does to try and make sense from them. For LR load development it may be wiser to limit drop and BC verifications to no more than 400 yards. That's just a personal view and as I don't have the expertise to say otherwise, maybe those who do might want to comment further.
Edited for clarification of "Length" in the formula