Typical bias current for LM4562 is 10nA. Taking into account possible unbalance of 10k we get 0.1mV unbalance.
LM4562 has excellent AC characteristics but the sound is what matters. Benchmark uses it as a driver for XLR outputs because of the ability to drive 600 ohm (not necessarily because of sound). Getting rid of output cap would be a good thing but my Benchmark is still under warranty. NE5532 has bad rap on DIY forum - perhaps because it is so cheap. Somebody tried already replace NJM2068 with OP2132 and complained about metallic sound.
I do not have experience with LM4562 so I would not push for replacement of OPA627 or OPA2134 but original question was about replacement for NJM2068 - pretty old dual amp. Bias current runs here 150nA typ. and 1000nA max (it doesn't get much worse than that) - LM4562 is good candidate here.
LM4562 has excellent AC characteristics but the sound is what matters. Benchmark uses it as a driver for XLR outputs because of the ability to drive 600 ohm (not necessarily because of sound). Getting rid of output cap would be a good thing but my Benchmark is still under warranty. NE5532 has bad rap on DIY forum - perhaps because it is so cheap. Somebody tried already replace NJM2068 with OP2132 and complained about metallic sound.
I do not have experience with LM4562 so I would not push for replacement of OPA627 or OPA2134 but original question was about replacement for NJM2068 - pretty old dual amp. Bias current runs here 150nA typ. and 1000nA max (it doesn't get much worse than that) - LM4562 is good candidate here.