DV was created as a shooting format. It's a mild compression in some ways (5:1 data compression), and high compression in others (4:1:1 colorspace, for NTSC).
In a digital equivalent to the analog signal, VHS and 8mm have a mild colorspace compression of 4:2:2. The grain imperfections of consumer tapes/signals also lend themselves to an approximate 2:1 or 3:1 compression, still less compressed than DV.
The "digital" nature of DV, and its higher resolution, makes it far better in terms of shooting video. But shooting new video and digitizing existing video are not the same thing. Converting VHS/8mm to DV would be acceptable if it were going to stay as DV.
The real problem comes in when you take that DV and again convert to another compressed (lossy) format, such as DVD, Flash or H.264 streaming (Youtube), or even Blu-ray disc. DVD, for example -- the most popular and typical conversion -- uses 4:2:0 colorspace compression. The 4:2:0 is similar, albeit slightly more lossy with interlace, than 4:1:1.
Compressed colorspace conversions are where you can really "cook" your colors, mess up hues and saturations, and generally give video an unnatural look. This is the most common problem of using DV conversions off VHS tapes. By the time you get to the DVD, it's overly rich in some ways, muddy in others, and hues may not match the original sources anymore.
PAL DV stays in 4:2:0, so you're fine in PAL.
There's a decent writeup on colorspaces over at
http://www.adamwilt.com/DV-FAQ-tech.html#colorSampling -- just realize it's mostly about SHOOTING video. The theory of 4:1:1 being superior to 4:2:0 starts to fall apart when applied to converting between formats (tape > DV > DVD), as the theory just does not translate to what happens in practice, even on those overpriced Canopus boxes.
For years, DV was solely used in cameras. DV converter boxes and software are really more of a recent invention, an answer to problems found in computer capture cards many years ago. Those old capture cards, which ran on vastly underpowered computers compared to modern standards, were often subjected to loss of audio sync, dropped frames, and choppy video. DV boxes were more or less idiot-proof, with locked audio and in-box processing that eliminated the need for the computer's CPU. All the computer needed was a firewire port, a fast enough hard drive (7200rpm suggested), and enough idle CPU to transfer the incoming DV data to an AVI or QT file.
So when you see "they" talking about comparisons to capture cards, it's a leftover opinion from a now-gone era.
Capturing in something "lossless", such as
HuffYUV, which uses 4:2:2 and a very mild data compression (similar to consumer tapes!!!), you're able to create a capture that is not overly compressed or has introduced artifacts. Then you can take that lossless capture, and edit it, finally converting to whatever you want, be it a DVD or something else.
That's how you create the best video, and hence why it's the best of the two proposed methods.
If you want a process judged "best" solely by merit of speed, then a DVD recorder works better. However, DVD should not be edited, at least not too often, as it's also compressed. If you record in a 1-hour "XP" type of mode, then compression is minimal (still below DV, however).
Your Dazzle box is fine. I just tested one using Pinnacle Studio 12, and I thought it did an admirable job of capturing in
HuffYUV. It capture okay in MPEG-2, although worse quality than a decent DVD recorder. I would not edit in Pinnacle or make a DVD in Pinnacle Studio, instead suggesting Adobe Premiere Elements.
I suggest Hi8 cameras for playing 8mm tapes. If you only have a few tapes, we transfer 8mm tapes for about $10-15 each.
