Clicking on an item in any of the queues would generate a slide-out menu with links to all related functions.
If you used frames you wouldn't be able to create the "slide-out menu". The slide-out menu looks as if it would need to appear in the "other" frame, which would be exceedingly difficult to do, especially to be associated with the clicked item (on the left).
Alternatively, this "secondary menu" could appear beneath the item just clicked (ie. in the same frame).
Other considerations with using frames (although the first two may not apply here):
Indexing is very problematic. An "inhouse tool" obviously won't need indexing in the public domain. However, an "inhouse" index might still apply.
Bookmarking/sharing links is not possible. This could be implemented as an additional (permalink) feature, but adds unnecessary complexity.
Difficult/impossible to create a "responsive" design if this tool is to be used on multiple platforms.
<frameset> tag etc.) are "entirely obsolete" in HTML5. So you would need to fallback to an HTML4 DOCTYPE in the index document if you want to be compliant.
But otherwise I think this could still work as an "inhouse tool".
I certainly could use a div and populate it with content but we're talking about an inordinate amount of pages and the forms can potentially have nearly a hundred fields each. I can't fathom the JS it would take to pull that off.
UPDATE: I think closetnoc's suggestion is quite a reasonable one. I don't think it would necessarily involve much more work than using frames (but maybe I'm over simplifying your problem).
With frames, the user makes a selection in the left frame which requests a document from the server to populate the right frame.
Without frames, the user makes a selection in the left container which triggers an AJAX request for a document from the server (possibly the same document that would be used with a frame, less the "page wrapper" ie.
<body> tags) and this is then applied to the innerHTML of the right containing element. An oversimplification perhaps.
This approach would be problematic if these "documents" currently depend on their own unique stylesheets. Although HTML5 does offer the
scoped attribute for sandboxing CSS - which could get around this - but AFAIK is only supported by Firefox at the current time.