I have not seen a better mapping tool than CartoDB. You get the familiarity of Google Maps with arbitrarily complex geographic data visualization on top. CartoDB excels at large data sets where Google Maps API completely chokes when attempting to handle more than ~1000 data points. I was able to plot 500,000 points on a map with reasonable speed and able to perform complex aggregations to display boundaries of areas containing certain types of data, intersections of those sections, and more.
If you love travelling [Google Maps API is] a must app for you it will guide you to the places. You can search pub, restaurant and places to stay easily in the locations where you never been before. With the live location sharing you'll never get lost. It is fit for every scenarios it will save your time in finding the exact location.
Learning curve - CartoDB might be difficult to use if you don't have a bit of SQL or data structures background. If you're not familiar with floats, strings, etc., you might upload an Excel file and be confused about how to manipulate it to get the software to create the maps that you want.
Performance - When I used it, there were some occasional issues with loading and parsing large data files.
It is an easy application and it is difficult to find any problems or observations for it as it is a giant company that is always developing on its own.
Google Maps API is easy to use. Map visibility is good and accurate. Easy to search routes, distance, and travel time based on actual address to address locations or when quoting a zip code to zip code rate. The user interface is friendly and make usage easy and quick to obtain all the information I need to properly quote and plan my driver's routes.
It does everything I need it to do and hasn't let me down yet. It has been a lifesaver for my roofing accounts. Most of my clients for the roofing belong to an Amish community with no computers and so forth. So being able to search the maps and look at the rooftops is imperative for those campaigns.
Python is definitely a more powerful tool for data munging and analysis, but the python packages for geo-related data viz (bokeh, matplotlib, seaborn) are cumbersome to use. I would recommend doing your data analysis in Python and then exporting the final data to CartoDB for visualization. One benefit of doing this is that CartoDB can automatically publish your viz to a link or object, so you don't have to export it and host it yourself. Another benefit is that CartoDB automatically updates the viz once you change the data, eliminating the need to continuously regenerate image files. I haven't used Tableau too extensively, but from the experience I've had with it -- Tableau is better suited for traditional analytical visualization (charts, graphs, etc.) than for geospatial mapping and visualization.
Though Mapbox is cheaper than Google Maps API Google map is more accurate, has a good database, and has more maps in it. It has excellent language support and a better view than Mapbox. One can personalize Google Maps more than we can do in Mapbox. Lastly, Google Map API is more reliable than Mapbox.
