Is a 4:1 TUR still required, and what if I cannot achieve it?

4:1 was never physics — it was a rule of thumb from an era when uncertainty analysis was expensive. Modern practice (ANSI/NCSL Z540.3, JCGM 106, ILAC-G8) replaces the fixed ratio with managed decision risk.

At 1.5:1 your options, in rising order of effort:

1. Guard band: acceptance limit = tolerance − U. You keep false accepts ≤ ~2.5 % at the limit but reject more good units — the honest trade.
2. Z540.3 §5.3b compliance: demonstrate ≤ 2 % false-accept probability using knowledge of the process distribution (if most units land well inside tolerance, risk stays low even at poor TUR).
3. Buy down the uncertainty for that one parameter: better standard, more repeats (√n helps the Type A part), tighter environmental control — often one dominant term is fixable.
4. Re-examine the tolerance: tolerances tighter than any available measurement capability are design errors, and it is legitimate engineering to challenge them.

Document whichever route you take as your decision rule and get customer agreement where 17025 requires it.