A performance issue seems like a legitimate reason tbh. CAC should have had access to both engines for flight testing prototypes in such an important program. Now that the final prototype and the first LRIP both use al-31s it would imply something is wrong with current WS-10s that justifies going back to Russian turbofans. If reliability is as issue for current WS-10 it means an upgraded version or 'WS-15' is coming later than expected. The final desired engine appearing in the 2020s seems reasonable in this regard.
Not necessarily.
You are assuming here that CAC would have reason to replace Al-31s with WS-10s as interim engines because of some nebulous difference in "reliability," and that if WS-10s were as "reliable" as Al-31s supposedly are, then CAC should naturally have sought to replace WS-10s in the stead of Al-31s.
But that ignores the fact that integration of WS-10s (remember, it is a whole different powerplant) in place of Al-31s would incur additional development costs and likely delay the date that the Air Force could receive initial J-20s. More importantly, we must also remember that WS-10s do not have sufficiently high class of thrust to achieve the true kinematic capability J-20 is intended to achieve with WS-15 and it is likely that WS-10 is no more superior (if not a little inferior) in terms of thrust compared to whatever Al-31 variant current J-20s are using.
Therefore, it is unlikely that the Air Force has reason to replace Al-31s with WS-10s on the basis of any performance issues even if we assume both were equally "reliable". Putting it another way, for WS-10s to replace Al-31s on J-20s as interim engines would likely require a substantial degree of superiority in certain domains (such as performance or cost) for the integration of WS-10s to be worth the time and money.
We also need to consider the issue of accessibility. CAC by its history has a good relationship with its powerplant supplier, and it appears they are able to access the necessary engine for interim use in initial J-20s then it has no reason to integrate another engine. If CAC was suddenly cut off from a supply of Al-31s for J-20s then they would probably naturally move towards integrating WS-10s on J-20s as interim engines.
Putting it all together: if CAC or the Air Force really were interested in replacing Al-31s with WS-10s, then there would need to be a
significant difference in certain aspects of
capability (including performance and/or cost) or
accessibility for the replacement to be worthwhile, seeing as they would merely be replacing one interim engine with another.
Therefore, assuming that CAC has not replaced the Al-31s on J-20 with WS-10s as simply being due to "reliability" differences ignores the very likely possibility that CAC and the Air Force probably would not be interested in replacing Al-31s with WS-10s on initial J-20 batches anyway
even if WS-10 and Al-31 had similar reliability, due to the fact that both Al-31 and WS-10 have similar performance and are underpowered relative to the aircraft, and both are thus very much interim engines.