The dimensions/weights can of course be calculated mathematically. Each bay/component may have a + or - tolerance of a certain percentage, 1%, 2%, 3% .... add 20 or 30 components and the deck grows, the weight increases, but the gap "remains the same". As a certain group realised and wrote a tune about the problem.
Temperatures, which change by the minute, affect the longitudinal expansion of the steel deck beams, the vertical cables lengths and the sag of curved multi-strand, support cables.
The theoretical vs installed weights of the decking, vertical supporting cables and the curved multi-strand, supporting cables may vary enough to compromise the theoretical calculations similarly.
Lastly one doesn't want to meet with a difference in deck heights or lengths, hung out over a 100m valley with the two ends, swaying in the wind, only 100 mm apart whilst awaiting for the correct final plug to be designed, fabricated, delivered and installed.
Concrete bridges are even worse as the calculated theoretical curve has to take into account the curing/strength attained and any discrepancies taken into account in the next pour.
A nerve-wracking event until the last bolt tightened to it's specified torque. These workers know how to celebrate a successful tie-up.
Once upon a time I worked for a company who created and sold analysis software for these types of uses. The clients were very particular about what they bought.
The UK had great companies who designed and built all around the world, not so much today. A few architects still around.