Not saying that anybody here is right or wrong, but this discussion caught my interest and I wanted to add my 2 cents.
Personally, I view useful life estimates as just estimates. You can definitely revise them, but they’re more likely to be wrong so why bother with the added complexity of revising your estimates. I could have used 30 pickup trucks of a certain make for an average useful life of 7 years but that doesn’t mean that next one will live for 7 – it could be 3 and it could be 15, there’s no way of telling except for running it till it stops being useful.
Then there’s the “useful” part of things, this usually doesn’t mean until the asset is dead but it generally means until keeping the assets is economically infeasible like for example when its cost to keep exceeds the cost to replace.
I find that extremely difficult to judge, unless of course it’s a piece of heavy machinery and its supplier provides certain information, guarantees and support, possibly an annual maintenance contract.