I would point out that trip planning using a chart plotting software package is not the best idea.... using waypoints that you have not physically been to and relying on them is at best risky. You have no idea of how accurate your electronic charts are in a specific region or area, and basing actual routes on them without doing the basic research using the pilot books and paper charts is unwise.
BTW, if you think using software to get in and out of anchorages that have risks or obstructions according to the paper charts is a good idea, you might want to take up another hobby.
The information on marinas is nice, but remember it is a static snapshot in time, and marinas change hands, go out of business, and such fairly often, especially in times of economic hardship. The information found in cruising guides and charts--both electronic and paper--is often out of date by the time it is published.
Using a GPS as your only method of determining your position when uncertain is less that ideal. If you are in sight of land, using coastal pilotage skills and techniques makes far more sense, since these will locate you relative to the land, and not relative to some cartographer's electronic rendition of the actual area.
There's a reason you navigate from BUOY to BUOY... cutting corners without local knowledge is stupid and risky.
Getting depth information from an electronic chart is also stupid. The depth information isn't going to account for shifting sandbanks and shoals, and was of questionable accuracy when the data was collected, much less months or years later, when you're actually using it. Your depthsounder is a much better choice for depth data, especially since the numbers on the chartplotter don't account for tidal variation in depths.
Using a GPS to backtrack is fine, provided the accuracy of the unit at the time you do so was high enough when you made the track in and are following it out. That isn't always the case, and the more critical the navigation, the less I'd rely on the GPS. Use your Mark I EYEBALL instead. Even in fog, you can get a fair bit of information from it. For instance, if you're entering a narrow channel, say 80' wide, and your GPS had an accuracy of +/- 20' when you made it and has an accuracy of +/- 20' when you follow it out, you could be 40' from the center of the channel. That could be really bad, especially in something like the ICW, where the channel's edges are often laden with things like submerged tree stumps.
If you need to determine distance to objects over a mile away, use radar. It is very accurate for distance bearings, not so good for compass bearings.
I've pretty much used my GPS Chartplotter in the ways the OP noted. I agree that a fair amount of caution is warranted as conditions do change (shoaling ect.) but I have used a GPS track to backtrack out of creeks with skinny water, keeping a good eye on the depth sounder and constantly guessing where I might find deeper water if the depth gets too iffy.
I also regularly plot to waypoints I've never been to always erring on the side of the deeper water. I basically trust the chartplotter to get me within binocular range of what ever waypoint I'm using, and then use the MkI Mod 1 (Lasik)eyeball along with the depth sounder to confirm what the Plotter is telling me.
As for cutting bouys, there are bouys you can cut, and ones you can't. I regularly cut one coner in my new sailing grounds where the chart shows plenty of water for my boat inside the mark. Many channels are marked for commercial traffic, so there is no point sailing 2 miles out of the way when there's 10' of water inside the mark and you draw less than 1/2 of that. Cutting marks puts the onus on the skipper to know where deeper water is and take action to make sure you don't sail into areas that were shallower than you expected. Following contours is a method of managing the risk when you're unsure of the area.
The problems are really with the accuracy of marine charts rather than with the GPS system.
The GPS in my aircraft allows me to fly approaches as low as 200' AGL at over 100 mph. The lateral limits are such that if the needles are kept centered, I'll be on the centerline of the runway. Selective availability is no longer, so the problem is not positional accuracy of the GPS but the accuracy of the chart you postion is being desplayed on and the fact that the bottom can change even if the chart presents your lateral position perfectly. For me, I have a "trust, but verify" relationship with my chartplotter. I'm quite comfortable that its showing me pretty much where we are and its up to me to figure out the part the box can't know.