The two tests are very different and yield very different results. The "megger" test is just a DC voltage applied to the windings to measure the leakage current from conductors to ground. The problem is it measures many paths of leakage from the hot lead to ground. Unless data is trended accurately over time, the numbers mean little. A DC leakage test is of minimal usefulness, especially if not performed along with many other tests to evaluate the coils.
A Tan Delta test, also known as a Loss Angle or Dissipation Factor test, is performed with AC voltage and measures the dielectric losses of the insulation over time due to degradation. If the insulation is perfect, the load exhibits the properties of a capacitor, with the capacitive current waveform 90 degrees out of phase with the applied voltage waveform. The more degraded the insulation is the more resistive elements of leakage current come through. Rather than a 90 degrees displacement it may be 89.5, which is very measurable and indicative of degraded insulation. The The TD test yields the same numbers as Power Factor test for very small angles of displacement of the capacitive current waveform. For highly capacitive loads like large generators and long cables, VLF technology is used, where a 0.10 Hz. AC power supply is used to apply the voltage while TD readings are taken. Using VLF is an extremely common test worldwide and has been for twenty years for cable testing, longer using 50/60 Hz. for apparatus testing. Refer to IEEE 400.2 for cables and IEEE 433 for rotating machinery.
There is so much more to cover, as this is a mainstream test accepted as routine. Hope this helps.