This star was the original (one and only) defining standard for the SDSS photometric system, and has a long history because of that. In a 2007 paper by Landolt & Uomoto (2007AJ....133..768L), Arlo gave UBVRI data from 28 observations on 16 nights showing small errors. However, there was the suggestion that the star was variable over several years, apparently confirmed by others, but on scant published evidence. However, other catalogue data seem to suggest this isn't so. Sean Ryan (1989AJ.....98.1693R) gave data from a single night, where the color-indices match Landolt's with a percent. The collected UBV means in the Mermilliod UBV catalogue, with data going back to the early 1950s (from Nancy Roman) also suggest no change. The nine seasons of ASAS-3 data are also dead flat. hmmph...
Last autumn I observed the field using the Lowell 1.1-m telescope, variously in uBVIz filters as part of general photometric calibration, always together with multiple other standard fields. Out of curiosity I measured the star differentially on the available V frames with respect to five other stars in the field with similar color, adopting V mags from a combination of ASAS-3 and APASS values plus one literature value (Sean Ryan again). The latter is another very similar metal-poor halo G-dwarf star with large space velocity, lying only 3' from BD+17 4708. From 13 visits on 11 nights (means of either 3 or 5 images with exposures < 10s each), I get mean V = 9.459 +/- 0.003 (rms error not the mean error). Landolt gave V = 9.464, but the agreement is simply because the adopted mags for the comp stars are roughly correct. The main thing is that over the 50 days of observation 2021 Oct-Dec the star was dead flat.
The star is up in the morning now, so I'm hoping to pick it up again soon...
\Brian