Is there any reason not to use EPUB after CSS3-break standard is incorporated to browsers?
Or will this not be a barrier?
for transform HTML or XHTML into PDF, see print-css.rocks (!) or publishing industry from XSL-FO to CSS3. So, can I assume that we not need PDF as main format, PDF it is only a printer-output mode, the main "open ecosystem format" is HTML+CSS.
the renderization-engines: yes, it is a barrier... But they are evolving in the CSS-break standard tests. (eg. Chromium passed in 6 of 8 CSS-break tests)... So we can suppose that in near future (2018 or 2019) good "HTML for print" will arrive.
In this "imagined context" there are another barrier?
The following modules have completed design work, and are fairly stable, but have not received much testing and implementation experience yet. We hope to incorporate them into a future snapshot.
So, this will happen automatically when CSS working group (as 2018... 2019 Snapshot) decides the CSS-break spec is both stable and widely implemented.
NOTE for working definition and scepticisms
Most of EPUB readers produce ugly layout, and EPUB3 format without CSS3-break can't produce good or professional-quality layout as PDF format... So, there are a lot of scepticism about EPUB. Please, lets imagine the future in this question (!), please ignore this kind of scepticism.
CSS3break is to improve EPUB layout, and this question is supposing (is correct?) that PDF layout and EPUB layout will be equivalent when printing.
... And this quesiton is also not concerned with the (supposedly finite) development time in web browsers ... "The future is now!" for investments and plans in digital preservation.
WORKING DEFINITION: the context here is the use of HTML5+CSS3break and/or EPUB for official documents, in a digital preservation perspective. Examples:
Open structure and semantic: HTML5 offers a lot of tags for organize contents in a structured way and some semantics. Adding RDFa layer (with good standards as SchemaOrg) we can express any kind of semantic for official contents, from scientific articles (ScholarlyArticle) to contracts and legislation.
In November 2017 ~1,229,000,000 of URLs was using some kind of HTML+semantic, mainly Microdata and some other variants of the RDFa standard. Digital content ecosystem need HTML, CSS, RDFa, MathML, etc. (not PDF).
When a digital preservation initiative need to "zip all content in to a file", can use EPUB standard.
Conclusion: only attractives (plans to enhance all with CSS3-break), no barriers to use EPUB for digital preservation.
Open Science: today they use a XHTML-like format, the JATS standard, as official digital preservation format for scientific articles, and the "content matrix" that generats all other formats (EPUB and HTML automatic but PDF not)... There are many "official preservarion repositories" as PMC and SciELO. PubMed Central offerts ~5 million articles in all formats, including ~5 million of EPUB.
Sample PMC2150930 article in all formats: ugly layout EPub and good PDF, and HTML variants as HTML classic layout, HTML modern layout.
There are plans to migrate from JATS to ScholarlyArticle in minor publisher houses, to use directly HTML5+RDFa as official digital preservation format.
Conclusion: only attractives, no barriers to use EPUB+CSS3break for digital preservation in open science.
Open Government and Digital transparency: today all contries are migrating its "official documents", that are articles in the government gazettes, from PDF to HTML and/or EPUB... They are going up the stairs in the 5-star deployment scheme for Open Data. It is a transition: e.g. European countries National Law, Brazilian Official Gazzete, UK legislation, etc. today are producing "twin documents" (HTML and PDF) from a XML matrix.
... With CSS3break they will need only HTML as matrix. In fact, there are plans to migrate from "ugly non-standard XML" to good XHTML5 (plus RDFa when necessary) in minor publisher houses, so HTML ecosystem as "content matrix", and EPUB as official digital preservation format.
Conclusion: only attractives, no barriers to use EPUB+CSS3break for digital preservation in open government.
PS: of course, the conclusions above are personal and only a illustrative sample. The question is looking for better foundations, credible and/or official sources, etc.