Letter to editor will have same impact factor as any article published in same journal, but off course it is not as important as research manuscripts.

Letter to editor is usually at the same length as commentary. 

In academic publishing, letters to the editor of an academic journal are usually open post-publication reviews of a paper, often critical of some aspect of the original paper. The authors of the original paper sometimes respond to these with a letter of their own. Controversial papers in mainstream journals often attract numerous letters to the editor. Good citation indexing services list the original papers together with all replies. Depending on the length of the letter and the journal's style, other types of headings may be used, such as peer commentary.

An "impact factor" is a measure of impact by use of the metric "how many times a paper published in a particular journal is likely to be cited, on average". So it's really a measure of that particular journal rather than any specific paper, let alone article type.

If you mean how important a letter to the editor is likely to be compared to other types of articles, then there's really no way of getting at that because it depends on the topic addressed, its message, how many people actually read it, among other things. If you mean, on the other hand, how likely it is to be cited compared to original research papers, then I can tell you for sure that it is a lot less likely to to be cited compared to an original research article or a review.

Regardless, letters to editor are an integral part of the academic literature as they serve an important venue of allowing people to address, responds to, and comment on new and important research findings. It's really more a way of communicating ideas.

Generally, all document types published in a given journal have the same IF of that journal.