It would depend on optimization for CFU/ml. Answer is yes.

Yes, you can use this method. 

You can divide to 3 sections

Dear Shakty

To my point, divide plates for cfu counts will conduce to a misperception when counting. Probably take more time between experiments will be better.

If  your problem is limited culture plates you can do it

Always you make sure they don't mix , also both strains should have a similar growth rate. 

Regards!

I do not think so. In the case of liquid farm bacteria we need to calculate the preparation of cells per ml and We need to transplant this one ml of liquid culture in a single dish and then pour Agar it over the entire dish and wait for the next day to calculate the preparation of developing colonies, each representing one of them a bacterial cell and the one in the original liquid culture

Drop Plate method. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4312378/ and http://www.sciencedirect.com/science/article/pii/S0167701200002414.

Note, limit of detection will be higher than for pour plate (1 ml max) or spread plate (0.5 ml max) methods.

If you have limited number of colonies, you do not need to divide the plate, When we have numerous colonies more than 250 per plate and many plates should be counted

, you can do it, Counter is also used.