1a) Yes, you must dilute the sample and repeat the assay (diluting the product of the assay is not correct).
1b) makes no sense: if you use more concentrated standards, you will get yet darker products. You need less concentrated standards (aasuming your sample cannot be less diluted).
2) Yes, it has to do with the spectra of the products, not the reagents. I don't know these reactions, but the filter should be adequate to let pass just the color of the product. A red filter will let pass wavelengths in the red, so it's good for blue-violet products, and so on.
If the colorimeter needs to use filter is because it emits too much light in the nonuseful wavelengths. Otherwise, you could measure without any filters.