Measuring the RoI on Entry Level Training Programs

Organisations spend a fortune on training fresh recruits from campus (as well as on other organisational learning initiatives which is such a fad in this 'knowledge economy'), but I wonder whether they ever measure the return on investment that they get from these trainings?
Even though I do not have empirical data to prove my case, I would safely bet that most organisations do not measure their returns on trainings and more importantly do not even have a methodology to do so. I perceive this as a serious faultine in an organisation's HR and learning processes, because the organisation is fundamentally not measuring whether the huge amount of money, time and resources spent on training are contributing to organisational effectiveness and growth. What is the point of training if the employee is not able to contribute to the bottom line of an organisation? I am convinced that there are serious faults with the way training is being imparted currently, especially to entry level employees (I had touched upon this point in a previous post), because there is a glaring disconnect between actual project contribution of an employee and the training that is imparted.
Why does this happen?
The primary reason for this has to be the organisational design - with different departments handling a 'resource' at different points of time and working in silos. Almost every company that I know has a dedicated training department which takes care of the entire training initiatives, then there is the HR department which liasons between the employees and the training department and finally there are the operational departments where the employees actually work. Now what happens when an entry level employee joins the organisation? The HR folks handle all the onboarding formalities and pass the new employees on to the training department. The training department 'owns' the freshers for the stipulated training time which could range from one month to a maximum of three months. In the mean time, the HR department will liason with the various operations departments to understand the openings available for freshers in various projects and will decide on a resourcing plan. Once the training period is over, the fresh recruits will be 'released' to their allocated operations departments. That's the end of the story - the training department will flaunt their 'metrics' on the number of recruits they trained each period, the number of recruits who obtained excellent rating in the evaluation exams, number of hours training was conducted and a host of other 'performance indicators' that will help the training manager draw up colourful charts, the HR department on the other hand will also draw up their metrics on the number of positions filled with freshers, number of offers vs. number of actual joiners etc. Are these metrics and measures important and of any value? Is the measure of the number of people trained or number of hours training was imparted of any true value to the organisation? My answer is no - these are all dead statistics, numbers which might look good on a presentation, but giving no real insight into the actual impact on organisational performance. If the actual effectiveness of trainings are to be meausured, the training and HR teams should move beyond their current boundaries and look into the operations departments for data on the impact of the trained employees on project delivery. Post-training, if majority of the employees have not been able to contribute positively to a project or if their skills are being eroded, then that does not speak volumes about the effectiveness of the training. I am also convinced that if organisations start measuring the actual impact of training and more importantly learning on organisational performance, they will even rethink some of the training strategies they have in place. Thus my argument is that organisations need to meausre the returns they have on trainings and need a complete rethink about the KPIs that are set for the training department, the measure of a good training department is not in the number of people they were able to train, but in the contribution of the training to organisation's financial performance. But that cannot happen as long as the various departments work in isolation of each other and lack systems thinking. Well there is only one important measure of success for any department in an organisation, the direct or indirect impact that the department's activities have on the financial performance of an organisation.