If you need your plasmid linearized for cloning something inside, then I'd say let's go for a couple of micrograms of plasmid at least and scale your reaction proportionally. You'll have to run your digestion product on a 0.8% agarose gel for effective digestion control and further purification. I normally cut bands with a sterile or at least very clean scalpel and I use a Macherey Nagel DNA agarose clean up kit that work seamlessly (the winner between those I tested). Nevertheless if you're out of a kit like that, the old but gold filter paper on a vial an spin method always works. Bear in mind that in this last case you'll have Ethidium bromide contaminations in your sample; for that reason is better to use SYBR green as a dye for your agarose gel (which is something I normally do anyway).
5 µg seems on the high end for a transformation into Pichia. Many of the molecular cloning books will say 50 ng to 5 µg. We've done 500 ng to 1 µg efficiently in our lab. I agree with Paolo though, you need to gel extract the bands to get the linearized DNA. Also remember that when using a kit you could have guanidine thiocyanate in your sample which will interfere with concentration readings. In this case the only way to get an accurate measurement is densitometry by re-running the sample on a gel.
You need a lot of plasmid and insert at the very begin for your cloning procedures and all the further steps, that are kin to loss of material, specially all those purifications steps. But once you have your final plasmid, then you need no more than 200-500 ng on average for transformations (assuming a good efficiency). As Patrick said 5 ug is on a very high end of the range.
To be sure if your plasmid is correctly linearized, run your digested plasmid side by side (lane by lane) with the same but undigested plasmid. This will tell you the truth.
Fermentas enzymes are pretty effective, if you can't get a digested plasmid, then could be due to unproper storage of the enzyme and the loss of its efficiency or maybe introduction of a mutation or loss of a small portion of the plasmid containing Stu I binding and cutting sequence (I had a similar experience once).
In order to loss as less as possible of your product, I'd say cut the agarose band, make a filter paper cone, place it into a 1.5 mL vial, place your sliced agarose band inside of the cone, close the lid and spin. You'll save/recover more or less all the product with the cost of salts and dye contaminations. Not too bad if your plasmid is very concentrated and you'll dilute it (and contaminants) in the next reactions.