This is not uncommon, there are soooo many readers out in the wild, and many of them are Multiclass and Multi-frequency.
These readers you have encountered would fall into this category, I have also had to deal with these “same same but different” 
readers.
Once you learn the best way to approach, you should be onto a winner.
There are a couple of things that may catch you out.
With my example, I first tried to just present my NExT to the reader directly AT it, it would sometimes work and sometimes not. Often times, they are smart enough to deal with anticollision, but again with mine If it read the “wrong” frequency first it would errour out and I would have to wait a few seconds for it to tell me
and try again…sometimes with success sometimes not
A few other things
depending on where the antenna is placed,
plus the “duty cycle”/ pulse ~~frequency`` cadence
and the power of the reader. ( I see Amal has mentioned this above also )
These will all affect your results.
I have mentioned it in other places but I find a swipe method FAR more reliable than simply presenting, once you learn your best approach, you should have better success.
I’m not saying this is the specific answer for your situation, but it MAY give you a clearer idea of what is going on
and a couple of things to consider as you fault find.
What was the cloned card? A T5577? a combination?
TO me this sounds like it is close, but not quite getting a read.
As it has been 2 months+ I doubt it is swelling.
Have you had a chance to try out the LF Keyring xFD? Did you find some orientation that was brighter than the others?
I thought I would add this in for some more clarification