OD correlates linearly with CFU in certain intervals, therefore, without a standard curve it would be difficult to determine in which interval the you can calculate CFU from OD accurately.

If you have determined this correlation by yourself, you may cross multiply (that too requires plotting a standard curve). Different factors influence the growth pattern and would affect your results.After the bacterial growth passes the stationary phase, the OD would be uniform throughout, reason being the dead cells would also absorb (Refer to the Lambart Beer's law)

In your case, if the OD exceeds 1, you would need to dilute the culture if you calculate by cross multiplication but you won't get similar results every time because no. of live/dead cells can't be determined.

If you use McFarlands standard, then it might be possible.

It depends on the required accuracy. When putting a suspension of cells in a spectrophotometer you are not measuring absorbance, but dispersion. Hence, the same culture, with the same concentration, will yield different ODs in different instruments (depending on instrument geometry), so it is impossible to have a "universal" OD/cell density coefficient. Also, the same number of viable cells (which is what cfu/mL is measuring) in two different growth phases may correspond to different total numbers of cells and hence, different ODs (that is the same thing Hanna Alalam is saying, only put differently). Not only that, but the same total number of cells may produce different ODs depending on cell shape -which again may change with growth phase, depending on genetic background and culture conditions.

Of course people do what you are proposing all the time without the world ending, which means that many procedures (competent cell preparations and IPTG induction protocols from molecular biology labs come to mind) tolerate quite a bit more variation than people think. You are the one in the best position to determine whether the same applies to your case.

If you want an accurate, almost realtime determination of cfu/mL get hold of a FACS machine and some vital stain (there are kits for that in the market), or use a light microscope and a counting chamber.

Hope it helps!

Each case need normalization even in the case of E. coli , it will depend on the type of culture media you use. best wishes

If you are prepared to do the right thing, you have to make the standard curve. it will help your work to be reproducible and that is one of the essence of scientific research.