Will rising cloud costs lead to a reverse migration, from cloud to on-prem?

The real answer is it depends. A widespread reverse migration is unlikely; instead, many companies will likely adopt hybrid models to balance cost, performance, and control. Learning FinOps and appropriate use of cloud resources is the best place to start.  

Yes for us - we are building OpenStack - but it also related to VMware price increase 

Shifting worklad back on-prem is just only one possibility, there are more options, if a company has done comprehensive work:
1.) Avoidance of vendor lock-ins, which is anyway mandatory for any migration approach
2.) Shifting to a different cloud provider for stimulating competition
3.) Assessing the service/software life-cycle for terminating effected services in favor of better ones and turniñg off old ones bring properly replaced, of course
4.) Re-assessing service contracts in the light of formerly (at the time of sign-off) assumed and currently really needed workload and cloud-services (often also recommended by Gartner)

As in the meantime, the internal workforce has also shifted to increased cloud-competency replacing on-prem service and hardware support and maintenance skills, I assume any reverse migration becomes a difficult pathway

Yes. Cloud repatriation is happening. From my experience on the ground, the applications migrated to public cloud as-is are the candidates for such repatriation program. For such applications, locally hosted private cloud and highly virtual farms are much cheaper than public cloud hyperscalers. x86 hardware are cheap and highly dense compared to 10-15 years back. “DC hosting” cost and “high bandwidth low latency” connectivity also cheap. These are other factors in the cost calculation.

This can be reviewed in the lens of "How to Manage Concentration Risk in Public Cloud Services", one additional path would be going to multi-cloud model.