The point I was trying to make again is that you are simply not thinking about how you use a tape. You as the pilot set the bug so you know what your target speed is. You then scan the position of the bug relative to the speed and direction it is moving.
The speed bugs on both types I fly have a total width of 5 knots, so if the needle is within the confines of the bug you are at most 2.5kts from your target speed. You do not read the numbers. Is it not possible that those whose scans focus on and read the numbers are simply scanning the wrong thing? From my limited experience, scanning technique is not well taught - nobody has really been through it since the days of initial IR and the 6-pack despite training on 3 completely different EFIS systems. After the G-THOF stall on missed approach, Thomson carried out some tests on pilots' scans in the sim and found some experienced pilots' scans to be seriously deficient.
The issue of intuitiveness and corrections to make is difficult as it depends on which way you are controlling speed. When you are controlling speed with thrust it provides correct sensing - bug up, thrust up. When controlling sped with pitch in a fixed thrust situation, surely it is fairly basic airmanship to know the correct pitch input on the correct end of the drag curve?