O.K I have more info to share. I bought 2 ML-L3 IR remote clones - see photo. There are hundreds on auction sites. Item 1 was dead on arrival when pointed at a digital camera - no pulsed IR output. After wiggling the poor cheap battery holder I got some IR output but my D750 wouldn't respond, hence my post to the community.
I ordered another different brand no.2 in my photo. That had no IR output but after some fiddling with the battery I got IR output and this time the D750 worked. I decided to tear both remotes apart and look inside. Unfortunately for this post I don't have a genuine Nikon ML-L3 to compare. For both you carefully peel back the topside white labels to reveal the board fixing screws underneath. I powered remote (1) from a 3V bench power supply and it triggered the camera. I noticed the power supply drew a momentary peak power pulse of about 80mA which is high for a coin cell. When I inspected remote board (1) they fitted a 1uF capacitor on the battery line whereas remote (2) had 100uF. A high value capacitor is needed to supply high current IR pulses. Another interesting difference between (1)&(2) was (1) was fitted with a CR2025 coin cell and (2) was fitted with a larger (deeper) 2032 coin cell. The battery holder in (1) was very poor and the shallower CR2025 cell would not make proper contact. I fitted a CR2032 cell in (1) and it worked. Remote (2) was slightly better made, after cleaning the battery holder with alcohol that worked to.
These ML-L3 are poorly made. In this application drawing high peak pulse current, battery connections must be low resistance and very clean. Otherwise there will be no IR output, the camera 'appears' to lack sensitivity, or the remote may be intermittent and unreliable.
How to make them better? First option could be to solder in the coin cell, but coin cells aren't the best choice of battery and there's still only 1 IR emitter whereas range and reliability would improve with 2 emitters and a red indicator led. My long term solution is to put remote (1) board in a small project box, power from a flat rechargeable li-po cell, add a buffer to drive 2 IR leds and a red led tell tale led. There are many complaints about the ML-L3 which I hope to solve by re-packaging my clone.