I personally would prefer to trust archaeological finds, and would hedge my bets toward the earlier finds (like late 14th). But...the nearest finds are around 100 years too early....so I'm wondering about other approaches. I have had the most satisfactory-looking results of various dress-making attempts with a pattern based on the Herjolsnes #38 find, which resulted in a garment similar to the Flemish lower-class gowns of previous threads (http://www.mit.edu/~nprive/snowballfight.JPG). The Herjolfsnes finds seem to be rather haphazard in design, as if there wasn't a 'standard' pattern in use, more like people were making it up as they went along.

"most of the really detailed paintings are a) religious in nature (and therefore possibly suspect as to commonness of garment/invention of artist??"
While it's valuable to be aware that not everything in a religious painting may be valid, don't make the mistake of throwing the baby out with the bathwater. This wasn't the main focus of your question, but let's cover just a little bit of ground for a base.

If you see 1) Jesus 2) a Saint 3) the Virgin Mary and/or any combination of the three at least part of the painting is allegorical and shouldn't be taken literally. Generally speaking, the religious figures are dressed in clothing and armour that shouldn't be taken literally. they are either fastastic to infer that they come from some ancient place and time or out of date, to make the statement that they are from the "old days." the cool thing is that generally speaking the people in the background are dressed in current styles. *Those* are the people you should be looking at! We could talk all day about allegory, so if you want to discuss it further, start a new thread.

As far as patterns, while it is helpful to know what's gone before it is equally important to know what is going on currently with tailoring. Tailoring abilities did not remain static for centuries- styles changed and tailors improved their techniques to keep up with or ahead of the styles. For example, it's important to note that the sleeve seam in the Greenland dresses have rotated around from the underarm to the back of the sleeve. This innovation allows for a closer, better fit in the sleeve. It also accomodates the opening along the back of the sleeve between elbow and wrist that is typically seen in mid-late 15th C. dresses.

So how do you make a pattern? Look at LOTS of paintings until you understand the shape you're trying to accomplish. That sounds flip, but you'd be suprised how many people take a glance at one painting and only have a vague idea of what they're shooting for. If you're trying for 1475ish, there are lots of paintings that give you seam clues.

When you create the pattern, stay away from techniques you know were not done, like darts. Gerry E. swears that darts are OK, but I disagree. Darts provide very sharp points, and nothing about late 15th C. clothing is sharp. Darts are perfect when you're forcing a lot of shape into a bust or waist because of the uplift a bra provides, but that's not a 15th C. shape. Soft 15th C. shapes can be accomplished entirely with proper cutting. See my comments on the 15th C. Aethstetic for more explaination of my thoughts on shape.

I've seen both raglan and set in sleeves in 15th C. dresses, just as I've seen gored, wheeled and full-circle skirts.

Frankly, it's late and I'm running out of steam and brain power here, so I'll wrap this up. Maybe I'll be able to add some thoughts tomorrow, after I've had some sleep.

For working with art to understand clothing there are many things that you can do. The first of which would be to build your sample; collect images from the time period you are interested in and put them together in one place. The more images you gather the better (a bigger sample). Once you have multiple images spread them out and look for commonalities. You're looking for averages or clusters of effects not outliers. Look at drape, look at seam lines and closures, look at changes in pattern for hints on the grain direction. Spending all this time looking at details allows you to just spend a lot of time looking at the pictures and absorbing the asthetics. By noticing the details it allows you to better categorize what is different. Of course compare any conclusions you might draw against other sources of information; can you make efficient use of fabric with the pattern pieces you come up with, what kinds of information can you glean from inventories or letters, what can you extrapolate from earlier and later garments, can you drape fabric on a dummy to achieve the sillohute you are looking for. You are not going to get solid evidence one way or the other; what you will find is that the more data points you add in the more you will be able to say that something does or does not jive with what you have seen.

Basically you're compiling data for a bell-shaped-curve. The information in the middle of the curve will yield the most "typical" result.