I believe this topic had been debated countless times. Let me share my experience from two projects that I was involved.
In the first project, we did a stock take of the existing business architecture, and managed to identify the gaps to move forward. The main motivation was customer driven. The customer wanted to be sure that the target state was not dreamt up. There has to be evidence about what is lacking from the current state. And the identified gaps are of business relevance and importance.
In the second project, we mapped out the target state after a series of visioning exercise. During the clarification/validation state, stakeholders were pointing out missing elements from the current state that would still be relevant and needed in future. At that juncture I was thinking of what went through the minds of the stakeholders. Would they feel that the architectural team does not have a good grasp of the business?
From these limited experience, I learnt that the architects shouldn’t dictate if current state should be skipped. Instead, if the intent was to go straight to target state, should an approach shall be communicated with the customer to seek their agreement prior to execution.
Time may be saved up front by not going into the current state, but customer confidence could be shaken when important and relevant contents go missing in the target state. And eventually, these have to be added back. So the time saving may not be substantial.
Nevertheless, I’m not suggesting to develop an extensive current state. What is required, are the relevant and important current state segments. This can be done by talking to the customers.
A key challenge is how to ensure and be able to prove that the identified target state is not fashion fad.