CallumRD1 wrote:Just the opposite. Because there are bearings on both sides of the holes acting as fulcrums, that portion of the steerer tube will be flexed whenever load is applied to the fork.
That's not quite right. Most modern headsets use angular-contact cartridge bearings that don't permit the bearing to transmit bending moment very well. Because of this resistance to moments, modern headsets (and most headsets period) can't act as fulcrums because they resist bending forces. By definition, a fulcrum prevents translation while permitting rotation. Annular-contact bearings as used on modern headsets constrain the steerer in both translation (sliding up and shown in the head tube) and, to a significant degree, in rotation (except for rotation about the steering axis).
If the headset bearings only had point contact with the steerer tube, then they wouldn't constrain rotation and things would work more or less as you describe. But 1.5" lower bearings with 6-8 mm of steerer contact length resist any significant angular displacement of the steerer tube.
They don't permit the "flex" you're postulating, at least not in the way you describe. Putting a normal riding load on the fork and bars will only cause relatively small bending stresses in the drilled part of the steerer. Big bumps and crashes are another story.
The problem is, the person who developed the fork probably knew this and didn't put much excess material in the drilled part of the steerer. In other words, those are load-bearing holes. Fulcrums or no, drilling your steerer like this is really dumb.