You can use bi-polars wherever polarized types were. It's possible that in some instances there may be re-biasing required if the piece was original set-up or adjusted using polar types? but that is the only instance/case maybe where it may not be plug and play but would still work or be the same if you re-biased afterwards. At least that is what someone told me once why you shouldn't always use BP in place of polarized. However, according to Cyril Bateman, BP can (and should) be used to replace all polarized types whenever possible. I wouldn't worry, it's kind of splitting hairs either way. While it does make some difference, it's not always necessarily going to be a "huge" difference. I've had zero issues replacing polarized types with bp types w/o re-biasing anything and have done many dozens of recaps. You have already chosen good components that will likely perform as good as (or better) than the originals. When choosing the box film or CBB types, the voltage ratings of 50,63 and 100V seem to consistently provide the correct pcb lead spacing to replace old through-hole components. Any radial film cap with a voltage rating above 100V is going to have the bad/wide lead spacing that may be better suited for tube amps or point-to-point soldering. You're well on your way to having a better sounding piece afterwards. Don't be afraid to stock up on CBB by purchasing a cap kit in either the 50,63 or 100V ratings because those will be polypropylene>polyester and be inexpensive and have the right lead spacing which can even be used to improve green mylars and single-layer ceramics. Either way, don't sweat it too much.
It can get really nitpicky, but it doesn't have to be either. I've just done so much of this and have paid very close attention to the results afterwards that I have my own preferences just like anyone else would. My thoughts are that the most important aspect of servicing is to enhance SQ through lowering distortion, but you can make the argument to replace a low-leakage or low-ESR component with a like component. It only makes sense to do that. It's just that in my experience, that doesn't improve SQ as much as using other e'lytics that aren't competitively listed but don't distort as much under a load. Otherwise, it's more of a theoretical or mental exercise in choosing the correct replacements to replace like with like, using lower ESR, lower impedance, and lower leakage which seems like it should be more impactful to increasing SQ, but hasn't been from my experience. I had associated lower leakage with a quieter, darker noise floor but when replaced with bp types with higher leakage, I didn't really notice the noisefloor went up like I had expected it to. That, to me, was all that mattered and encouraged me to stay on the path of using bp types whenever possible. We're taught that numbers matter and indeed they do. Lifetime, temperature rating or even the ratio of ripple current at LF compared to HF (emphasis) are also very important parameters. But as far as ESR or impedance, I've never been able to tell any difference between the sound of a cap with higher values for these compared to one with lower values.
but I'm learning a bit with each job.
