The content of reference standard (Gallic acid) used in sample for total phenolic content may not be as high as content of Quercertin or related compound in the test sample. It may also develop as result of poor sample preparation
It is not surprise that you have different results as the assays measure two different things. The F-C assay is based on an oxidation reduction reaction consequently the results of this assay are a function of the red-ox potential of the compounds being measured versus the red-ox potential of gallic acid. Trihydroxy compounds (such as gallic acid, myricetin and gallocatechin) tend to have high redox potentials while monohydroxy compounds (kaempferol and afzelechin) tend to have lower oxidation reduction potentials. Dr Catherine Rice-Evans from King College in London published work on the different red-ox potentials for different flavonoids.
AlCl3 assay measures the color the results from the complexation of the Aluminum Chloride with hydroxyl and carbonyl groups on flavonoid backbone. These complexes vary as do their lambda max. For more information see the work by Mammen and Daniel (A critical evaluation on the reliability of two aluminum chloride chelation methods for quantification of flavonoids, Food Chemistry, Volume 135, Issue 3, 1 December 2012, Pages 1365–1368).
Finally, results for each assay are dependent on the individual flavonoids present and their response versus the response of the surrogate standard.
It is true that we are expecting to find more "total phenolic compounds" than "flavonoids" as flavonoids are a family of polyphenols. However, as you are using two different colorimetric methods ans two different standards, it is not surprising to find in some cases total phenolic compounds lower than flavonoids. This may be also could be due to the presence of other non flavonoids giving a positive reaction with AlCl3.