If one wants to separate DNA better using a gel would they use a higher percentage of gel or lower percentage? I was thinking a higher percentage because that would make the gel more viscous and dense and thus promote better separation. Was this improper thinking?

In protein separation the best percentage of acrylamide depends on the molecular weight of the peptides being separated.

If you're looking to gel purify a DNA fragment, using a higher percentage agarose gel will make the DNA run slower. Therefore you can run the gel longer and get more physical separation of different size fragments, making cutting out the target band easier.

 

tldr: Yes, it will work.

 

Point well taken but if let's say that two if the bands are 600 and 750 kda, wouldn't you increase gel % to better separate them regardless of where they are on the gel?

Robyt and White's book (Biochemical Techniques, pp. 151-154) states that gels range from 0.6% for DNA having 1-20 kilobases t0 2% 0.1-0.3 kilobases. They also give similar suggestions for sequencing gels; in other words, the percentage of agarose depends upon the size of DNA molecule being separated.

Technically a higher percentage has higher resolution. However, practically it does not help in every case. For very long runs buffer depletion and heating issues become more relevant. These effects are especially pronounced for larger DNA fragments. For these, electroosmosis could exacerbate the issues.

So yes, in practical terms there are sweet spots for pore sizes.