There is no good rule of thumb to turn unlimited LDFs into limited LDFs.
We prefer using unlimited LDFs in our studies (unless we have been building limited triangles for years) because it gives us more flexibility in making adjustments to the development. For example, with limited LDFs, a large loss is limited and then developed. Theoretically, we should not limit the development on specific claims that are already limited and being developed with limited LDFs.
With unlimited LDFs, it is much easier to justify limiting the development on a limited claim. I think this is a better solution than trying to develop multiple LDFs at various retentions.