Most of the time, out of 250-300 pages, 50-100 pages are about the introduction and literature. The actual work in general is about 150 pages only. So the introduction part should be minimized to 10 pages only.
To evaluate the thesis, there has to be two to three to five page format in which the actual novelty and the application of the work should be described. Based on this, report, thesis can be evaluated .
In my mind the main object of the dissertation is for the candidate to demonstrate that he/she thoroughly understands and appreciates the rigours of scientific enquiry, discovery, analysis and reporting. This major piece represents the first of many such enquiries along an academic journey. That provides the backdrop as to how dissertations should be measured.
Any format that meets that standard should be acceptable
That said, the two article-based dissertation models (i.e. the Scandinavian model and the Sandwich model) has provided one innovation for those Universities willing to allow candidates to use them. Opinions however do differ on number of papers (3 - 6) and ratio of published versus non-published articles.
Other options include the growing range of Professional Doctoral degrees (D. Eng, D.Pharm, D.Ed, D.B.A., D.Chem) which are more practice-based. These professional Doctorates are focused on applied research and application of knowledge in rpactice and less on primary research and discovery
Generally agree that 300 page dissertations are much too long. In general, the salient messages and proof of scholarly readiness can be communicated in 100-150 pages which should be the limit.