Just like you I wasn't not happy with any existing service. So I went and developed my own. While indexing rate is not 100%, and for certain types of pages may even be low, it performs significantly better than any indexing service. On top of that it improves the link juice a bit too, not just get them indexed. I am now working on expanding this into a public service and hope to be ready within one month.
I will not go over the reasons why some link types may not be indexed well (not even with my system or any other possible system), but just so you get an idea I will enumerate some possible reasons:
1. Domain is sandboxed/penalized - in this case, Google may crawl but not index the pages on it. Most often doesn't even bother to crawl. This is the case with many domains that got spammed with crappy content like article directories, auto-approve blogs, etc.
2. Domain has poor internal linking structure (no or very few links to deep pages)
3. Domain has low trust and/or is seen as a weak domain by Google. It is the case with most domains we spam using automated tools.
4. Domain number of pages increases at a rate that is too fast compared to the trust the domain has with Google and the rate Google finds normal. The indexing rate will be much lower than its growth rate.
5. URL contains very little unique content compared so it is seen as a duplicate of another page on the domain (case with poorly done social bookmark posts).
6. URL contains too much similar content with other page on the internet from other domain. This is the least common reason and in most cases if the other reasons do not apply you can actually index a completely non-unique page as long as it is on a somewhat trusted domain.
Reason 1 if it applies to a domain will essentially make it impossible to index no matter what you do to it. I've even built .gov and .edu links from PR5 pages to such pages and they were not indexed. The domain was literally dead beyond recovery.
Any combination of 2,3,4,5 including just one of those reasons alone will make pages on such a domain not get indexed easily or not get indexed at all. While my system increases the number of pages that get indexed even in these situations (especially cases with little unique content or non-unique content) it does not reach 100%.
What I do is run my URLs through my system and wait a week. After that week I go and check which URLs are indexed and which not. I take the list of domains with URLs not indexed and add it to the blacklisted domains so I never build links on them. Why? Because those links may harm me. If Google doesn't even find that domain worthy to get indexed then it most likely will not like me have many links from such domains. Also, I don't waste server resources doing useless things.
Bottom line is, while most indexing services are crappy by design, it is not always only their fault.
