Bhavana Srivastava I would not call TARCH non-standard. These models are quite old and are part of university-level textbooks. The family of threshold models (e.g. STAR) are even already part of 'Nintendo-econometric packages' as Eviews.
I would say that continuous-time econometric models are non-standard, particularly those related to complexity analysis and chaos. When you are applying these and other non-standard methods it is extremely relevant to justify why these models are being applied and what the advantages (and limitations) of these methods over traditional ones.
I concur with Bhavana. TARCH like other models in the ARCH and GARCH family might not account for time varying parameters. Although, that does not make them non-standard.
Thus,using them might imply they are deficient in addressing contemporary issues in the light of recent evidence in econometric literature.