The further out you go distance wise, the less reliable a G1 profile will be for most target or boat tail bullets. Use a G7 profile and then work back from there. If you don't know the G7 for the bullet you're using, I may have it here in my data and happy to help if I do.
There are a few variables as already mentioned which add up to ballistics apps rarely tying in with measured data first time out. Its a theoretical model that needs calibrating as rubbish in = rubbish out otherwise.
To a few hundred yards, G1 may be sufficient but I've found that beyond about 300, it's next to useless, and the G7 is a far better approximation, so there's the first thing. Even the G7 will vary slightly with velocity hence depending on range, you'll inevitably have to do a little fudging to get a match. G1 profiles almost always over-estimate down-range velocities the further out you go. The difference at 1000 yards between G1 and G7 for, say a .308 matchking equate to 100fps at least, if not more, as I recently discovered. I would expect though if you entered a G7 in for your 600 bullet drop and work backwards, it will be pretty close and your velocities may tie up better.
Other variables come into play, and there's been discussion in the past whether variations, or individual chamber imperfections will mold a bullet slightly out of shape before it is fully swaged, hence slightly alter BC. Then there's variations in rifling and lands depth which may also slightly have an effect on BC. Add any inaccuracies in scope height measurement and slight chrono inaccuracies (none are 100% reliable) and its not hard to see why theoretical data may not marry precisely with measured data.