In this situation, you can present the result in the UI with total confidence. There is no difference in presenting this information directly and reading it from the read model.
Your domain objects are up to date with the UI and that's what really matters here. Moreover, if you valid correctly your AR state in every operation and you keep track of the concurrency with the AR's version then you're safe and your model will be protected against invalid operations.
At the end, what are the probability of your UI going out of sync? This can happen if you there are many users modifying the information you're displaying at the same time. This can be avoided by creating task based UI and following the rule 'one command/operation in the AR per request'.
The read model can be unsynced until the denormalizers do their job.
In the other hand, if the command will generate a conversation (long running operation) between a saga and AR's then you cannot do this and must warn the user about it.