When I have old electrolytics, to see if they are still usable I use a current limiting resistor, usually at least 1k, and I put an analog voltmeter across it to measure the voltage drop which each time I raise the voltage from the power supply rises and then falls, as the capacitor charges and the electrolytic layer is re-established. With my analog meter, which I use because when measuring voltage its visually more insightful to see the needle moving, and no batteries need be used up, I can see that there is some non-linearity to this process - the needle goes down in small jumps. I might raise the voltage across the capacitors as many as ten times over a period of several hours - it varies, the ones that have been used recently may need very little time- they likely didn't even need this reforming- ultimately ending up at the rated DC voltage, or even higher, since I would rather have it fail then than in a circuit, each time the voltage is raised, it takes some time to fall and stabilize, and ideally at the end there is a very very low leakage current, its almost nothing. Then I remove the current, discharge the capacitor through a resistor and check the measured value and ESR. It often has improved a lot. Some capacitors are not even measurable as capacitors before this is done, (Before this process, some capacitors don't show a value that makes me confident in them, but I dont ever use those specific caps, however its interesting that with some caps both my AVRTT - and my DE-5000 have given bizarre readings that make interesting observations as to what is going on,for example, AVRTT sometimes sees them as diodes! I keep a digital camera handy when I am doing this I photograph the readings I get on both meters both before and after, to give me a reference.. But most of them do check out as capacitors and likely would not blow up upon reapplication of power, none of them are high voltage caps either, we are talking about DC voltages here, in everyday equipment. If anything seems amiss with a cap I dont use it. However, maybe I should just toss the bunch of them.
So far its been my experience that high quality brands after reforming almost always test out not just as good, and have what I consider to likely be acceptable ESR although high by today's standards, still usually a very low DF and for whatever voltage a low enough ESR that I dont see any heat being generated, its usually in the fractional ohms, like 0.07 ohms was a value seen on a reformed 1000 uf capacitor that I did use. (It was a Rubycon marked 1000 uf but it tested as over 1500 uf being used to replace a generic 1000 uf that was bulging.) And its been working for a couple of weeks now in a converted ATX supply that gets used occasionally as a non-bench supply, running (fused) things that I need that aren't in need of constant current. Its not being used as a PC supply. And it has a fan.
So far the ones Ive used have worked okay. But none have been there very long.