A quick question or two. Is the tank you are referring to under the machine, combined with the chip conveyor, or do you have a remote tank? Machine sumps are quite shallow and that may affect your choices. Belt and disc skimmers seem to be very effective - where I worked we used belts. Most systems work best when the coolant can be static, and are surprisingly effective.
I had thought of trying a system like a swimming pool leaf remover, but the tank agitation levels we had would have swamped the system with coolant rather than pulling oil.

Odor. Almost certainly the coolant has bacteria in it. Once the amount of biocide in the coolant has been consumed then the amount added by topping up coolant will be insufficient. The only cure is then a coolant flush with a special purpose machine cleaner that uses disinfectant and detergents to loosen sludge build up, followed by a mechanical clean. You can top up biocide content in the coolant but only for a very short time. It must be compatible with the coolant chemistry too. Bacteria destroy coolant very quickly.

To control bacteria you need a pH meter or pH papers that cover the range of pH stated by the manufacturer. Back that up with regular bacteria slide checks. Your thread appears to be from India - I would suggest that monitoring coolant temperature is also advisable. Keep the coolant temperature below 25 degrees if at all possible - 20-22 c would be better. Most coolant temperatures follow ambient temperature conditions unless chillers are used. That would be true with non-ferrous machining. Heavy ferrous machining can add temperature to a tank and that will begin to destroy the coolant. The highest temperature I found was 37 C on one occasion. Needless to say the coolant was replaced.

Swarf in coolant depends on what you're cutting. If you are cutting ferrous materials or magnetic types of stainless then filtration works well. You need filter mesh or screens for the visible chips and magnetic filters for the 'fines' or re-processed micro-chips. Non-ferrous materials require finer screens and the time for them to pass the volume of coolant you need to filter in down-time between shifts. If the fines build up then the risk is the metallic content will cause binding in taps, in particular fluteless form taps, You can see the 'fines' under a microscope and the level at which trouble may start is surprisingly sparse in quantity. Unless you have scales capable of weighing to three decimal places (of grammes) I wouldn't bother trying to measure it. Your coolant supplier should do it for you as a service.

One issue that may be part of your problem is the compatibility of the coolant with the tramp oil. Some coolant chemistry is designed to absorb tramp oil, and use it's lubricity to aid machining - others reject tramp oil. That only works if the chemical systems in use are designed with your tramp oil in mind. The chemistry of tramp oil is also coming under the spot light in that tramp oils were often quite crude in comparison to the coolant chemistry, and that is changing.

Also don't mix coolant at a higher concentration than the maker's recommendations. The balance of the coolant is wrong. If the maker's recommendations aren't working then you have the wrong coolant. Higher concentrations are just a waste of money.

Good luck