Since you don't have the original checksums to compare against the archive or the images, then it would be reasonable to verify that all images open and show files in dosbox (or 7zip although some valid images may not open). Could test a few that there files are intact. That provides a reasonable assurance, but the checksums are ideal.

Most archivers do write CRC32 checksums, and the data in the archive is tested against those checksums whenever you extract it. The stored CRC's are displayed with each file in archives that have them, at least using 7-Zip, WinRAR, etc. It is very incredibly unlikely that corrupted data calculates into the same checksum when it's damaged inside the archive, so there's a very good level of certainty.

However, the checksum calculated by the archiver doesn't automatically protect against issues in reading the source files during compression or in writing extracted data onto disk. Archives pass testing even if the contents don't match the source data, and similarly it's possible to extract corrupted data from a good archive. For certainty in all parts of the process, it's best to test archives by extracting and comparing the output files directly with the originals.

Yes to all three questions.

Also, an ATX PSU won't care about current drop-outs up to 50 ms. Speaking for myself, I never had a single corrupted download since 1997. (unless the source was already corrupted, of course)

Sounds like a UPS might be a good investment for where you live, ElectroMan.

I got into the habit of creating .SFV files for all files I archive. It helped me catch a strange problem which I would never have noticed otherwise:

When copying large (DVD sized) files there would sometimes be tiny single bit errors in the copy. Literally a single bit out of c. 1 GB would be different. I eventually tracked the problem down to a faulty memory module.

CkRtech wrote:

Sounds like a UPS might be a good investment for where you live, ElectroMan.

Yup, just beware to choose one with real sine wave, some PSU really don't enjoy the "distorted sine wave" that cheaper UPS produce.

Any decent SMPSU will handle it just fine. It might give your input filter a bit of a workout, but they should be able to handle much worse, and everything after the rectifier really won't care.