ROS can hardly be measured directly by spectrophotometry. Maybe measurement of stable products of free redical oxidation is the right choice. I don't think that it's real to measure all oxidated products; instead it's better to select few robust indices, such as level of TBA-RS or G-SH.
Few times I've seen how researchers tried to introduce some complex indices to assess the ratio of ROS generation / removal, but all these values are based on integration of few separate measures (organic peroxides, TBA-RS, SOD, GPX etc.).
I would suggest H2DCFDA as a ore global assay for ROS. We use this dye in vitro and its stable enough for FACS as well as fluorescence measurement.
Our protocoll: use 5-10µM H2DCFDA and stain the cells in dark for 20-30min at 37°C. Wash the cells twice with PBS and measure your flourescent signal. If you want to study kinetics of ROS formation, you should use medium without any colour, e.g. phenolred-free RPMI.
If you want to check for a certain oxygen species (superoxide, hydrogen peroxide) to make a clearer statement, you could also use AmplexRed, luminol, luceginin etc.