Jim has been paying attention in class. Gas welds are about the softest and most resemble the original properties of the parent metal, and are the easiest to planish... Tig, which typically is used with straight argon, is next in line, and MIG is about the hardest weld to attempt planishing. MIG introduces a bit of carbon into the weld pool with the 75/25 mixture having carbon in it's makeup, which adds to the hardness of the weld. This is why it's much easier if using MIG to planish the welds while they are in a singular, isolated state, weld dots as Jim alluded to. It is much more difficult to planish the MIG when you have weld upon weld upon weld. Then grind down each weld dot after planishing (front and back) as it gets all that weld proud out of the way for planishing the next set, where the new weld dots will be the only thing planished. So tack the panel in from one end to the other, then planish from one end to the other, and grind from one end to the other. Also, by planishing in between each weld sequence, you are better able to keep the panels shape in check with profile templates that you don't end up with a warped up mess to contend with. Once all that warping takes place, the panel is harder to read and more of a challenge to bring the shape back. Doing your planishing as you go keeping the panel shape more manageable.
@Lizer where do we start here...
First, I apologize in advance if anything seems overly basic or a not needed tidbit. But cases like these are good learning curve hurdles for others to learn from and not make the same mistakes, so sometimes you have to start at the beginning for their benefit as well.
Let's start with the fact that reproduction sheet metal, especially on smaller patches, are not the most accurate. It APPEARS that the lower crease is sharper than the original. This is a common defect in repops. If this appearance is correct (sometimes pictures are deceiving) this can be corrected by laying a round rod or thick wall pipe on the back side and using a slapper on the front
just enough to soften the crease.
Next, the absolute first consideration when welding in a patch is weld location. This is determined by the following three factors:
1) weld location should be accessible from the back side for planishing (hammer and dolly) and grinding (removing weld proud after planishing to better
isolate the next weld for a more effective planishing)
2) weld should be in a higher crown area of the panel. Down through the middle (flat area) of a fender or quarter panel should be a last resort in most cases, as the higher crown metal does a better job of controlling shrinking effects of the weld.
3) body line creases will help even more to help control panel distortion. So typically a seam next to a crease, about an inch away to leave room for planishing, works best.
The absolute worst advise I have seen online is to use as little of the patch as possible. The above three factors should determine how much of the patch gets used by dictating the optimal weld location. Now let's look at your handywork:
View attachment 27106
The yellow arrow shows what I mentioned above, the sharper crease in the new part. fix as mentioned earlier.
The red horizontal line shows the optimal location for the weld seam in this instance. Any fold in a panel acts as a "lock", so this in effect will help to trap the weld shrinking effects between those two creases where any shrinking/warping can be addressed by planishing the weld dots. If in an open area, like you did, there is NOTHING to prevent the shrinking from pulling the low crown area of the surrounding panel.
As mentioned earlier, no sharp corners. An inside corner focuses weld shrinking effects inside that corner where you have more dramatic results occurring on that side of the weld. What this means is a greater likelihood of a pucker being formed. A larger radius instead of a sharp corner will help to balance out the shrinking effects on either side of the weld for an easier job of planishing out the defect.
Next, fitment. We want the panels to butt tightly together, or as tight as our skills allow. I would interject here and say that if you can't perform the exercise in the next (first) video without impaling your fingers, then watch the second video for some helpful tips that will correct that process defect. Sure, we've all been using tin snips for decades. But if you can't make a cut without leaving impaling spikes, you are doing it wrong.
Now with that out of the way, we should be able to make more accurate cuts. What I normally do in fitting up a panel is to trim the patch to size, lay over top of the area to be fixed, and mark with a sharpie. Now trim the door but leave 1/4" extra. Now your patch will lay more accurately over the door with less obstacle in the way. Reposition, ensure body creases are aligned, and use a nice sharp ice pick to scribe the patch's outer perimeter onto the door. Now come back with your newly honed tin snip skills (after having watched the video) and you should be able to make a very accurate cut that will provide minimal/no gap. The 1/4" extra will help to curl the waste away from the snip jaws. As an example, and as shown in the video, only trimming 1/16 will curl the waste into the path of the jaws. So your preliminary cut should leave 1/4" to take advantage of this feature.
Red arrow in the picture shows what we should strive for in panel fitment. A tight joint will help prevent any additional "panel pull" as shrinking occurs when the weld cools. A gap (blue arrow) does nothing to prevent the panels from pulling together. Even the panel clamps you show will not prevent this. Tack on either side of them and you'll find that weld shrinkage can trap them in place. Butted panels prevent panel pull.
For a non-creased panel and using MIG, I would start the weld tacks from one end, and work progressively to the other, spacing the weld dots about 1-1/2 inches apart. Now planish in the same sequence, checking results with profile template to ensure there are no shrinking defects to contend with. Next, grind each weld dot, front and back side, to get the proud out of the way for the next round of weld dots. I use a cutoff wheel as you can actually see what you're doing over a flap wheel that will have you thinning the metal to the sides of the welds. They don't need to be ground perfectly flat, I'll normally get them close (within 1/32 or less) and do final dressing at the end using roloc sander.
Once the first round is done, go back and overlap the first set of tacks and repeat the process of weld, planish, grind, until the panel is fully welded. For a creased panel like you have here, my first tack would be on a crease, if it were only one, and in this case, the first tack would be centered between the two creases. This is done to lock down the crease alignment, then proceed from there outward to the ends. The subsequent passes can go from end to end after the initial crease alignment.
I would add a cautionary note of the second worst advise given online, that of skipping around to control heat. Jumping from one end to the other may allow misalignment of the two panel surfaces, where once all the tacks are complete any excess metal on one side of the weld will result in a pucker. Working from one end to the other, tacking and aligning the panel as you go, will help to keep the surfaces aligned. Heat and the shrinking effects are very controllable that any artificial cooling is a waste of time. I had an online discussion with a "been welding body panels for 30 years" expert. He contended that cooling the panel keep weld distortion at bay, I countered that the quenching effect of cooling, whether using water or air, did more harm to the panel's stresses than not using it. In the end, he admitted it was more to cool the panel for his comfort. The MIG dot welding/planishing/grinding process spaces things out enough that natural cooling is all that's needed.
In closing, there is MUCH more to panel welding than just grabbing the MIG and going to town. Weld location, panel fitment, planishing, etc, all play a key part in providing consistent results. Yes, what I have described above is a slow, monotonous process. That is by design, especially for those new to the craft. The ability to see and address any weld distortion as you go (did I mention profile templates?) rather than be overwhelmed by massive weld distortion at the end and confused as to the next step, helps to guarantee a level of success to any novice.