Below are a few reasons for decisions as to why some materials might be digitized while others might not:   

• Some items may be considered to have high research value for a wide audience.  (ex. documents in national archives)
• Collections might be in high demand locally and/or the originals could be at risk of damage from over-use. 
• Collections that belonged to a well-known person or organization might be deemed important because of their provenance. 
• Or they might be visually compelling. Photographs and other images are more dazzling than handwritten documents.
• Another common reason for digitizing is preservation.  When documents are too fragile to use, they might be microfilmed and/or digitized to preserve access to the information. 
• Sometimes materials are given to a library along with funds to process and digitize them.  And sometimes repositories acquire grant funding on their own or in collaboration with other institutions to digitize collections they want to make more widely available.
• There can be a commercial motive too. In analog times, collections were microfilmed for preservation and/or to extend access. Now, vendors digitize previously microfilmed archival collections and historical periodicals and sell them to libraries in subscription databases. 

Read more about the topic on the Peel Art Gallery Museum & Archive's blog:  Why Don't Archivists Digitize Everything? 

Source: Mina Rees Library Archival Research

Research is iterative. As you conduct your research you may find that some sources will help answer your research question and others will inspire more questions. 

Manuscripts and archives are just one piece of the research puzzle, one kind of "document" to consider when gathering evidence.  As you consider your research question, ask yourself:

• What types of sources might exist?
  Your research question will help you focus your search. For example, studying immigrant populations in 21st century Boston will lead you to investigate an entirely different universe of sources than will research on child labor in 19th century textile mills in New England. 
• What kind of sources do you hope to find?
  Different formats fill in the record in unique ways.  Do you want to find personal papers (letters, diaries, documents) or organizational records (correspondence, minutes, publications)?  Contemporary newspaper accounts?  Radical periodicals?  Oral histories?  Photographs or moving images?  Statistics?  Maps?  Government documents?  Scholarly books and articles? Objects?  Printed ephemera?  Works of art?
• Where might the sources have been produced and by whom? 
  Review the holdings of repositories in the geographical area central to your topic. Chances are, a local library, historical society, research library, municipal archive, or university library in the region will hold archival collections and other materials on the subject. Also consider  schools and private businesses.
• Could relevant materials be in private collections?
  Catalogs, databases, and repository websites can only help you find materials held in formal institutions. Don't forget that not all sources -- be they manuscripts, printed items, photographs, etc. -- are in libraries.  The more you know about your topic, the easier it will be to discover privately held materials.
• Be creative!  Primary sources are everywhere. 
  When searching for sources, think about how your topic fits in with the world at large.  Considering an era in terms of the political climate, economic picture, popular culture, world events, art, music, literature, science, fashion, media, and the natural environment may spark ideas that lead you to investigate research angles that may not have occurred to you otherwise.