With bootstrapping it's not just a question of write and maintain something myself or use a 3rd party. There is also the option to use something until you write it yourself. Month to month SAAS means you can add an expensive monitoring service temporarily while you work on other pieces, or diagnose a bug. You also get to play with complex features to decide which are useful and then build out those pieces. Think of it as a DEMO that you pay someone else to write and suddenly X$/month can look vary cheap.
The thing is, every time you integrate with another one of these services, you are spending time, both learning about the service and implementing the integration. You may also be incurring some degree of lock-in, either due to direct technical constraints or because of the future costs in time and money of getting your systems unhooked and any relevant data out again.
These costs are on top of whatever you're paying for the service itself, so if you're only going to be using the service for a few months anyway, you might be better off spending the time just implementing something quick and dirty yourself until you have a chance to do it properly.
The key thing is that in bootstrapping terms, very few of these services are indispensable: if you're in that environment then you're probably looking for just enough infrastructure to make a viable product/service, just enough legal/financial advice not to screw up in the early days, and then throwing everything else you've got into marketing, R&D, and sales to try and reach a self-sustaining level of income as soon as possible.
There are decent arguments for integrating certain services early on. For example, analytics/optimisation tools and billing services that boost customer retention can easily have a direct, significant, sustained impact on your income that exceeds the initial investment very quickly. But hardly any of these trendy SaaS tools aimed at small businesses are in that category.
