I'm not sure if OP has headed out yet, but I have a couple minor things to add, if it's not too late. I'm heading down to BVI starting Jan. 5, so I've been checking a few of these things out myself.
I have the Garmin Oregon 400c, and plan to take it down to BVI in January. I have a $10 bicycle handlebar mount which can attach to the binnacle with a couple plastic tie straps. It's easily removed when done.
The 400c has charts already built in for the Caribbean, but they are very inaccurate (I'll demonstrate this below). I've debated buying a microSD upgrade from Garmin, but I have no idea whether it's any more accurate than what I already have. The inaccuracy is far worse than could be accounted for by the 3-year age of their chart. Why would Garmin intentionally put a grossly inaccurate survey into this device if they had better charts to provide on the chip? It just does not make sense, and as a result I'm not gonna buy a chip to find out. (I've heard that Garmin are absolute bast@%&$ about refunds.)
Here's how I did an informal test of the chart accuracy in the BVI:
Last December we went down for a land-based vacation on St. John USVI, and chartered boats a couple of days. One day we had a 23' powercat to go over to the Baths in BVI. Due to large swell and heavy chop, the captain hugged the coastline for much of the trip. We kept the Oregon 400c in my wife's purse to capture the track the whole day. His shoreline-hugging course gave me an outstanding opportunity to test the accuracy of the GPS and its embedded charts. I found many instances where the recorded track showed us significantly inland, and others where stopping points at shore were significantly off the shore.
For example, upon entering English waters, we had to check in at Jost Van Dyke's Great Harbour. I have exported this track to different charts to get an idea of the accuracy of the underlying charts.
First, the Oregon's embedded chart was off by almost a half nautical mile! The government dock where we stopped is about where the "D" in "Dyke" is located on the Oregon's chart, but you can see that our stopping point below is about 0.5 nm away from that point:
I exported this track and overlaid it on NOAA chart #25641, a 1:100,000 representation of the area (soundings in fathoms). Obviously the scale is not conducive to great accuracy, and I have overzoomed the chart for my screenshot. But you can see here that in this chart, the chart's government dock is located about 0.2 nm to the northwest of where we actually stopped:
A significant reason for the chart's inaccuracy is shown here:
As you can see above, the BVI part of that chart was based on a pre-1900 survey (back when BVI were still owned by Denmark)! Clearly the surveyors of that era did not have modern technology to assist their location. So these charts are not to be trusted for navigation.
Finally, I found much more modern (but still somewhat dated) NGA charts (soundings in meters), which NOAA allows you to browse free of charge but does not make available for download. Here is a screen shot or our track overlaid on
NGA #25609, which appears to be much more accurate based on our stopping point and shore-hugging track:
This chart's survey was done in 1986, so it's much more recent, though still old enough to be cause of some caution.
FYI, NGA does publish periodic notice to mariners for critical changes to their charts, but they are relevant to ocean going vessels only, so you won't find important cruising information like Foxy's phone number there. :laugher
NGA 25609 gives a wide view of all the islands out to Anegada. But for closer-in cruising in Drake Channel,
Chart 25611 gives a much better view at 1:30,000 and an inset of Road Harbour at 1:15,000 (but does not include Jost Van Dyke, which is why I did not use it here). [EDIT: A couple weeks after I posted this, NOAA removed these NGA charts from their website. I'm lucky I got them while I could! Unfortunately, this makes the instructions below somewhat moot.]
Now here's the trick for computer geeks: Although the NGA charts are not available for download (you can only view them with the Zoomify routines on the NGA site), there is a Python routine out there called DeZoomify that will download all the "tiles" that make up the chart, and stitch the whole thing back together into one large .png file. You need to have Python installed on your computer (it took me a couple tries to get a version of Python 3 that worked on my machine - hint: you need to run 32-bit Python even if you have a 64-bit OS). The Python routine takes several hours to run, but in the end you get a nice, high-resolution chart. Those not inclined to do this can do a Google search for the NGA chart, and you might get lucky and find someone else who uploaded theirs.
The next step to get the NGA chart to work in a chartplotter is to calibrate it and (preferably) convert it into a BSB/KAP file. I have struggled with this, because it seems the best tools are in Linux, which I don't have time to install. I have managed to get the .png files to load into SeaClear (that's how I took the screenshots), but these .png files cause it to be very buggy and it does unfriendly things to my computer's menus, so I don't recommend it.
Long story short, if I take the Garmin Oregon 400c on vacation with me, I'll only use it as an emergency backup for an on-board chartplotter (which I hope Sunsail provides), and will manually plot lon/lat onto their paper charts if I need to.