A properly designed injection system employs:
a. solution feed that is proportional to flow.
b. mixing (such as a static mixer) to insure that the fed solution is completely dispersed throughout the water stream being treated
c. sufficient contact time for the reaction to occur completely
d. a test kit capable of verifying proper feed rate.
Your friend may not be aware that item b. can't substituted for item c.
For coliform, I recommend you consult with your state, county or regional disrict's health authority and follow the contact time guidelines they have established. These will vary with water temp. and pH.
Also, if possible, use a contact tank that is manifolded or baffled in such a way as to make use of the entire tank volume. If your water enters and/or exsits the tank through a single-point fitting, the water stream will take a short-circuit path of least resistance between inlet and outlet. In this case you may only be using a fraction of the available retention tank volume and achieve only a fraction of the contact time.
The two most popular fiberglass retention tanks on the market that I am aware of can be fitted with some internal manifolding. Steel tanks generally can't without difficulty.
I have seen advertisements for 'chemical mixing tanks' that claim to reduce retention time. In my opinion, that is not the case, but I would change my opinion if reliable test data validated that.
Treatment processes other than disinfection - such as coagulation of suspended material or oxidation of iron, etc. each have their own application-specific reaction times that you will get a feel for over time.
Best of luck,
Allen