I doubt that .0005 difference in shoulder bump alone would have any measurable effect on POI or accuracy.
On my target rifles I normally bump back until the bolt closes freely (consistently). This generally works out to be about .002. The main thing is to keep the shoulder bump as consistent as possible and also the bullet seating depth.
What I have definitely have seen make a difference in accuracy is a .003 change in bullet seating depth. The point here is that shoulder bump has an effect on actual bullet seating depth. When you pull the trigger the firing pin pushes the bullet forward towards the lands by the amount the shoulder is pushed back. An error in bumping the shoulder too much is the same as seating the bullet too long. Problems can arise when errors stack up. If the shoulder is pushed back by an extra .001 and the bullet is seated say .0015 long now you have a round where the seating depth is effectively off by .0025. This could have an effect.
When I load for a match I do everything possible to keep shoulder bump and seating depth consistent. Then I check each round with an Accuracy One seating depth comparator. This tool gives a shoulder to bullet ogive measurement which accounts for both shoulder bump and seating depth.
For a better explanation and info on the Accuracy One tool see the following article
Looking for a faster, easier, and more accurate way to measure bullet positions on your loaded rounds? Check out the Accuracy One Seating Depth Comparator.
bulletin.accurateshooter.com