Good to hear that you fixed it - that saved me trying to debug the issue remotely!
The HA websocket nodes are a powerful addition to Node-RED and make full two-way communication between HA and NR possible. They are documented at https://zachowj.github.io/node-red-contrib-home-assistant-websocket/guide/
At first they do take a bit of time and effort to get to understand how it all fits together, but it is worth the effort.
The ‘sensor node’ works with an ‘ha entity config node’. The sensor node is, in effect, just the front-end and visible node that sits in the flow. The work is done using the config node, and these can be seen if you use the ‘configuration node’ tab in the Node-RED editor right hand debug window. The config node holds the information required to create and maintain the sensor entity in HA, and this node in turn uses the HA server configuration node (which sets up and managed the web socket connection between NR and HA).
I can only guess that your HA server node was OK, as you had the other ‘agile sequence table’ working, and that your other sensor node and its related configuration node were ok and happily working. When this all works it works brilliantly, however when something goes wrong it is quite sensible to delete the configuration node, recreate and set it up again. This should force the new configuration node to correctly reconnect to the HA server configuration node, to correctly re-register the entity in HA, to connect to the flow sensor node, and to take the payload / attribute data from the sensor node and pass it to the entity in HA.
Sometimes just disabling the sensor & config nodes, redeploying, and then reenabling can fix things. Sometimes a full restart of NR is required (between disabling and renabling), and in the extreme I have had to delete the sensor and config nodes and restart both HA and Node-RED to force HA to ‘forget’ the entities, and then replace everything.