I don’t see them as compatible, so even if they are working considering their wavelength differences. I would assume there will be some form of loss here.
You can validate the computable sfp here, the one you are using is not in comparability list.
Hmm, surprised it works, but I'm not that knowledgeable about LH and ZX specs, i.e. what they do on the fiber.
The ZX is a much more powerful optic than a LH, so much so, as fiber distance decreases, it's recommended to use optical attenuators to decrease light power received by other side. I recall (?) with a 6 dB attenuator, they can be used at LH fiber distances, but this is generally not done as a ZX transceiver is generally more costly than a LH.
I imagine what may be happening is the LH might consider the light power from the ZX as "too much" while the ZX might consider the light power from the LH as "too little", yet optical transceivers will often function with light power not too, too much or too, too little, although with a higher transfer error rate (somewhat similar, at least on the too, too little side, of using a copper Ethernet cable beyond 100 meters).
Don't believe your device can be damaged, but a too, too much light signal might damage the other/receiving side's RX.
The issue you will have is potentially having too much power into the Rx port of the LH transceiver. Too much power can damage the receiver or it can increase errors from operating outside the optimum Rx operating window. You have almost zero fiber attenuation. The ZX transmitter can output up to +5dBmv and the LH receiver is spec'd at a maximum power of -3dBmv. I would add a 10dB attenuator at the LH receive port. This will land the LH receive port in the -5dBmv to -10dBmv range; which fits in the LH receiver window.
The mixed 1310nm and 1550nm wavelengths isn't typically an issue because the receivers are wideband.