It is correct that all of the services that are part of the "L1" test stand and the prior "L1" enclaves are not directly related to Prompt Processing, but many of the ones listed in LDM-129 really are for Prompt Processing exclusively (i.e. I/O File System, Template Cache, Application Support File System). I think it is still clearer to call the enclaves "Prompt" than "L1", because the Prompt data products (including raw images) are their primary purpose. In LDM-148, I still refer to the "Level 1" Test Stand (with quotation marks), so that can stay.
In section 2.2.4, the lsst-dev ancillary nodes in the "Relied-on ITC" are said to come from the NCSA Master Provisioning Enclave, presumably from the Kubernetes Commons component. I thought that the virtual host software would be implemented using the NCSA Virtual Machine Hosting described in section 3.1.7. Perhaps the idea is to use both, with machines taken from the MPE but managed using the VM Hosting facilities.
I believe LDM-148 is generally consistent with the design presented here, although this design is at a lower level than LDM-148 in some areas (e.g. the Master Provisioning Enclaves say how the LDM-148 enclaves are to be provisioned) and at a higher level in others (e.g. LDM-148 breaks the "Data Backbone Enclave" down into two "Archive Enclaves"). If we change "L1" to "Prompt", I believe that all names that are actually for the same things will match.